PORTFOLIO MANAGEMENT:

THEORY & PRACTICE

LAST REVISED
APRIL 2008

SCHULTZ COLLINS LAWSON CHAMBERS, INC.

INVESTMENT COUNSEL
455 MARKET STREET, SUITE 1450
SAN FRANCISCO, CA 94105
(415) 291-3000

www.schultzcollins.com
COPYRIGHT 2008

TA B L E O F C O N T E N T S

Preface: Obstacles To Prudent Investment Decision Making ...................................................... 1

Obstacles to Prudent Decision Making ................................................................................ 2
Chapter One: Basic Investment Concepts .................................................................................... 5
Investment Objectives and the Investment Policy Statement ............................................ 5
Investment Prudence............................................................................................................. 7
Market Efficiency ................................................................................................................... 8
Risk ....................................................................................................................................... 11
Diversification ...................................................................................................................... 12
Asset Allocation .................................................................................................................... 17
Chapter Two: Asset Classes And Asset Class Investing ............................................................. 21
U.S. Equities ......................................................................................................................... 21
U.S. Fixed Income ................................................................................................................ 24
International Equity.............................................................................................................. 26
International Bonds ............................................................................................................. 30
Real Estate ........................................................................................................................... 33
Emerging Markets ................................................................................................................ 37
Chapter 3: Dimensions of Risk and Return ................................................................................ 41
Aspects of the U.S. Money Management Industry ............................................................. 41
Value/Growth Asset Class Investing v. Undervalued Stock Picking ................................. 43
The Historical Evidence: Is Value a Strong Law of Asset Pricing?..................................... 46
The Small Company / Large Company Dimension ............................................................ 50
The Three Factor Model: Empirical Results........................................................................ 53
Factor Loading and Unsystematic risk ............................................................................... 56
Chapter Four: Building the Portfolio ............................................................................................ 58
The Risk / Return Continuum ............................................................................................. 58
Starting In ................................................................................................................................ 62
Defining Asset Class Weightings ......................................................................................... 63
Chapter Five: Investment Strategies and Investment Vehicles ................................................. 65
Active versus Passive Management ................................................................................... 66
Active Management ............................................................................................................. 66
Performance of Active Managers ........................................................................................ 70
Survivorship Bias ................................................................................................................. 73
Performance Consistency CASE STUDY: THE FORBES MAGAZINE HONOR ROLL .......... 74
Evaluating Active Manager Performance ........................................................................... 76
Passive Management .......................................................................................................... 79
Evaluating Passive Fund Performance ............................................................................... 80

Chapter Six: Portfolio Management............................................................................................. 82

Asset Allocation Approaches ............................................................................................... 82
Theoretical Payoffs to Different Asset Management Approaches .................................... 83
Impact of Trading Activity on Portfolio Returns .................................................................. 85
Liquidity Costs of Portfolio Management Strategies ......................................................... 87
Trading Decisions, Best Execution and Loss of Investor Wealth .................................... 88
Taxes, Inflation and Turnover .............................................................................................. 89
Conclusion: Independent Investment Counsel ........................................................................... 92
Investment Policy and the Prudent Investor Rule .............................................................. 93
Our Approach to Portfolio Supervision................................................................................ 93

P R E FAC E : O B STAC L E S TO P R U D E N T I N V E ST M E N T D E C I S I O N M A K I N G

More money is better than less. The paradox of investing is that, although investors generally agree with
this statement, the pursuit of more money is not always prudent. This paradox explains, in part, why Schultz
Collins Lawson Chambers, Inc. [SCLC} is not a money management firm. Money managers often seek to
generate attractive performance results by trying to identify undervalued securities with above average
prospects for future growth or income. Money managers market the portfolios formed from these mispriced
securities either to the retail public or to wealthy individuals qualifying as sophisticated investors under
current securities laws. The easiest way for a money manager to claim superior performance is to outperform
either their peer group of competitors or a benchmark such as the S&P 500 Stock Index.
While this may sound like a good idea, many investors lack a clear understanding of the functional
relationship between their personal and unique investment goals and the index returns that they see on the
nightly business report. Is the return of the index sufficient to fund their future consumption and wealth
accumulation objectives? Does the risk of an index align with their personal risk tolerance? Is the money
manager taking greater risk than the index? Despite, or, perhaps because of the difficulty of interpreting
personal goals in terms of risk and return, for many investors the investment problem reduces itself to finding
a money manager with a good track recorda manager who can beat the market. Curiously, however, a
money managers primary goal (the speculative objective of beating the market) is only tangentially related to
the investors objectivessecuring a retirement income, accumulating funds to pay college expenses,
maintaining wealth sufficient to make gifts or bequests, and so forth. Prudent investment decision making is
complex and extends well beyond the single dimension of historical track record.
Prudent investing requires that the risks and returns of the portfolio align with concrete investor objectives
rather than with abstract beat-the-market goals. Undoubtedly, portfolios must generate returns sufficient to
support the legitimate needs and expectations of their owners; however, such a portfolio is best synchronized
to the investors aspirations rather than designed to outperform a peer group. Investment strategies desiged
to maximize expected return may prove to be either financial bonanzas or financial catastrophes. Investment
strategies designed to enchance the probability that a critical goal will be successfully met, however, are more
prudent and suitable for most investors. Prudent investment decision making begins when the discourse
shifts from discussing how to maximize return to determining the risks and returns required to secure an
economic future. If you dont need to outperform the S&P 500 to have a secure economic future, why should
you take the risks necessary to do so? SCLC helps investors design portfolios to meet their savings and
consumption objectives, within the preferences and constraints imposed by their personal circumstances and
risk tolerance.
Historically, the professional U.S. money management industry has offered investors a treasure hunting
model. Success under the treasure hunting model is a function of the managers skills in selecting
undervalued securities and in timing price movements either between or within capital markets. The treasure
hunting model requires correct and highly concentrated bets. For a variety of reasons, however, this ability has
proved elusive. Although treasure hunting has produced examples of investment success, it has not served
the average investor well:
the industry looks very much like an unconcentrated, highly segmented, service-oriented
industry for which perceptions of the qualities of individual firms vary widely over time and
across customers. The structure of this industry is not unlike that of hair salons or trendy
restaurantsMoney managers who can provide a good story about their strategy have a
comparative advantage. In fact, the product sold by the professional money managers is not

just good performance but schmoozing, frequent discussion of investment strategies, and
other forms of hand holding.1
In the following pages, SCLC offers a view of portfolio design and management that is more prudent than
traditional treasure hunting. A prudent investment approach begins by identifying the returns required to
generate money sufficient to meet the wealth accumulation goals or cash flow liabilities that the portfolio must
discharge. Although returns above the risk-free rate require investors to take risk, investment risks must be
commensurante with the return objectives. Furthermore, both risk and return must be measurable and
consistent with investor needs and risk tolerance. A prudent approach evaluates the evolution of the portfolio
not solely in comparative terms (did I do better than a benchmark or a peer group?), but also in terms of
progress towards objectives.

OBSTACLES TO PRUDENT DECISION MAKING

Many people have difficulty making effective investment decisions. Investors face significant obstacles:
Complexity - informed financial decisions require insight into abstruse financial, economic,
and mathematical relationships; and may require serious introspection to define personal
objectives;
Uncertainty - decisions must be made without complete knowledge of future consequences.
Good decisions do not guarantee successful outcomes; bad decisions may result in outcomes
that succeed by mere chance;
Conflicting Objectives - an investment decision may facilitate progress towards one objective
(e.g., generating current income in support of a dependent) while, simultaneously, impeding
progress towards an equally important objective (e.g., wealth accumulation);
Lack of Perspective or Multiple Perspectives - issues may be difficult to resolve because
differing perspectives on the same data set can lead to different conclusions. An investor
who remembers the Great Depression may, upon reviewing stock market returns, reach
conclusions entirely different from those of an investor who started accumulating wealth in
the 1980s;
Information Overload - investors are deluged by an ever-deepening torrent of financial,
economic legal and tax information. Determining which data really matter may be close to
impossible.
Successful investors must overcome these obstacles. A general understanding of the fundamental nature
of capital markets and investments is a necessary starting point. Recent advances in the scientific
understanding of markets make it possible to deal with each of these obstacles systematically.
This document provides an overview of relevant academic developments pertaining to financial economics
and portfolio management. We first address some basic investment concepts that investors must consider in
formulating any successful investment program. These include investment prudence, market efficiency, risk,
diversification, and asset allocation. We then proceed to a more detailed discussion of the mechanics of
portfolio construction and wealth management. Topics include:
Characteristics of asset classes that may be included in a portfolio;
Determinants of return:
Designing a suitable portfolio structure;
Lakonishok, Josef, Shleifer, Andrei & Vishny, Robert W., The Structure and Performance of the Money Management
Industry, Brookings Papers On Economic Activity (Brookings Institution, 1992), pp. 339-391.

Selecting appropriate investment vehicles; and

An overview of alternative asset management approaches.
Often, there are no absolute right or wrong decisions. Depending on the context of the decision, one
course or another may be appropriate. As Independent Investment Counsel, SCLCs primary job is to
1. Help investors make good decisions by informing them about the merits of alternative investment
elections;
2. Facilitate the implementation and supervision of a portfolio once investment strategies and
tactics have been selected; and,
3. Evaluate the economic consequences of asset management decisions as the portfolio unfolds
over time.
This monograph introduces investors to important investment topics and to fundamental concepts in
asset management. It is not an introductory textbook in investment analysis; nor, is it a self-help guide on how
to pick winning securities. Rather, its purpose is to provide a non-technical understanding of
1. the nature and scope of prudent investment policy decision making (good decisions increase the
probability of good outcomes);
2. how to evaluate the variety of portfolio management elections that are available to you; and,
3. how to select the portfolio management elections that are best suited to your particular
investment objectives.
Investors sometimes confuse making money or generating investment returns with investment policy
my policy is to make money. However, a return generating process that does not flow from a carefully
considered investment plan often relies on mere luck. Depending on the nature and scope of the investment
endeavor, investment policy encompasses the planning steps required to determine:

Which assets are suitable investments (asset selection and retention policy)?

How should the investor combine the assets into a portfolio (asset allocation policy)?

What is the portfolio return required to achieve my investment objectives (required return vs.
desired return)?

What is the portfolios risk (the probability of failing to meet the required return over the
applicable planning horizon)?

Does the portfolio require tax payments (tax management policy)?

Does the portfolio require periodic adjustments to maintain its strategic asset allocation
(rebalancing policy)?

If the portfolio is funding periodic distributions, how much money can be spent without
jeopardizing the investment capital (distribution policy)?

If the portfolio consists of several taxable and non-taxable accounts, what is the optimal
location for the assets (asset location policy)?

As the portfolio evolves through time, how are investment results evaluated (monitoring and
evaluation policy)?

If this process is completed with sufficient care, skill and caution, the investor may begin the return
generating process armed with the confidence that investment choices have been well considered prior to
their implementation. Although a prudent decision making process cannot guarantee a good result, it

greatly enhances the probability of a successful financial outcome.

SCLCs website (www.schultzcollins.com) is a source for more information, often at a greater level of
detail. Previous issues of two publications -- Investment Quarterly and Fiduciary Forum -- may be found on the
website. Additionally, the website contains working papers, reprints of published articles, and links to other
helpful information and resources.
For clients of SCLC, there is a special password-protected area that permits access to course notes from
Economics 746: Asset Management. This course, taught by Patrick Collins, is part of the University of San
Franciscos Masters of Science (combined MBA / Masters in Financial Analysis) curriculum. The lecture notes
provide an advanced discussion of many of the topics covered in this document.

C H A P T E R O N E : B A S I C I N V E ST M E N T C O N C E P T S

Certain basic concepts are necessarily incorporated into any investment program, whether or not the
investor is aware of their significance:
Investment objectives and the Investment Policy Statement;
The Prudent Investor Rule;
Market efficiency;
Risk;
Diversification;
Asset allocation.
INVESTMENT OBJECTIVES AND THE INVESTMENT POLICY STATEMENT
Every investor has objectives, however loosely defined. When objectives are not clearly and consciously
articulated, investors may make decisions that actively frustrate their attainment. It is therefore wise to clarify
investment objectives, so as to gain a clear understanding of what the portfolio is intended to accomplish.
It is particularly important to distinguish between desired return and required return. Investors may
believe that high rates of expected return are better than lower rates of expected return (the nonsatiation
principle states that more money is always better than less). However, the connection between rates of
return and spendable dollars is opaque. Higher rates of return entail greater risk; and, the greater the risk, the
greater the uncertainty regarding future dollar wealth. Another way of understanding invesment risk, therefore,
is to treat it as the level of uncertainty that critical goals will, in fact, not be reached. This risk is known as
shortfall risk. Good decision making is possible only when the investor knows that the portfolios expected
return is sufficient to the task and that the shortfall risk is within the bounds of prudence.
Compound return, which determines the portfolios terminal wealth, is approximated, in lognormal
distributions,2 by the equation:A
Compound return = average return -1/2(variance of return).
This equation says that dollar wealth is increasing with respect to the portfolios rate of return, but
decreasing with respect to the variance in the return.3

A lognormal distribution has a bell curve shape.

spend dollars not rates of return. It is possible to construct a portfolio with a high expected rate of return that
produces a paltry amount of spendable dollars. This result is another investment paradox; and, is a reason for establishing
sound investment policy prior to buying or selling securities. As Yogi Berra remarked: If you dont know where you are
going, you may end up somewhere else.

3Investors

Mitigating varianceor, as it is usually expressed, standard deviation of returns4therefore makes the

wealth-building process more likely to succeed. In fact, an investor can often generate more long-term
spendable wealth by employing risk control strategies desiged to mitigate variance than by simply chasing
higher expected returns.
It is apparent that mere labels and slogans [e.g., growth, double-digit return, safety, low risk,
balanced, all-weather portfolio, disciplined investing process] are too subjective and ill-defined for
prudent portfolio design. It is folly to Implement the portfolio process without understanding both the
compound return required for savings and consumption objectives, and the likelihood of falling short of
portfolio objectives. In a nutshell, the portfolios expected return must align with the required return at a risk
level that allows for a good nights sleep. The prudent investor seeks not just expected return, but overall
welfare. Furthermore, the prudent investor monitors progress towards the goal (and makes appropriate
adaptations as the future unfolds) rather than trusting in blind luck.
Investors should document investment objectives in a written Investment Policy Statement [IPS]. An IPS is
the document that avoids ill-defined and subjective investment labels by operationalizing the key aspects of
portfolio design and implementation. The IPS takes verbal expressions of economic objectives (safe,
aggressive, etc.), translates them into quantifiable measures, and outlines the strategies that will promote
their successful attainment. An IPS separates the amateur investor from the professional investor. It
expresses the investment objectives unique to each investor, defines the strategy through which important
economic goals will be attained, and puts forth a system through which progress may be monitored and
measured. In this sense, investment policy comprises the set of guidelines and procedures that direct the
long-term management of a (portfolios) assets.5

4The

statistical definition of variance is the dispersion of the return generating density function about its mean. Standard
deviation is the square root of variance. The standard deviation statistic that tells you how tightly all the various examples
are clustered around the arithmetic mean in a series. The standard deviation of a series of asset returns is a common
measure of the volatility, or risk of the asset. When the data set are tightly bunched together, and the bell-shaped curve is
steep, the standard deviation is small. When the examples are spread apart, and the bell curve is relatively flat, that
indicates a relatively large standard deviation. All else equal, there is greater uncertainty regarding the final outcome.

One standard deviation away from the mean in either direction on the horizontal axis (the red area on the above graph)
accounts for roughly 68 percent of the samples in the set. Two standard deviations away from the mean (the red and green
areas) account for roughly 95 percent of the total sample set. Three standard deviations (the red, green and blue areas)
account for about 99 percent of the data points.
Bailey, Jeffrey V., Investment Policy: The Missing Link, Pension Fund Investment Management, ed. Frank J. Fabozzi
(Probus Publishing Co., 1990) p.13.
5

An IPS can ensure that a portfolio does not seek contradictory objectives from the outset. More important
in practice, however, and with a greater beneficial effect on long term returns, adherence to the procedures
set forth in a Policy Statement deters hasty, ill-considered reactions to market volatility.
The principal reason for articulating long-term investment policy explicitly and in writing is
to...protect the portfolio from ad hoc revisions of sound long-term policy, and to...hold to longterm policy when short-term exigencies are most distressing and the policy is most in doubt.
History teaches that both investment managers and clients need help if they are to hold
successfully to the discipline of long-term commitments. This means restraining themselves
from reacting inappropriately to disconcerting short-term data and keeping themselves from
taking those unwise actions that seem so obvious and urgent to optimists at market highs
and to pessimists at market lows. The best shield for long-term policies against the
outrageous attacks of acute short-term data and distress are knowledge and understanding
committed to writing. All too often, investment policy is both vague and implicit, left to be
resolved only in haste, when unusually distressing market conditions are putting the
pressure on and when it is all too easy to make the wrong decision at the wrong time for the
wrong reasons.6

INVESTMENT PRUDENCE
Whether or not investors define their investment objectives and procedures in a written Investment Policy
Statement, they enhance their chances for achieving a successful outcome if they adopt a prudent investment
approach. But what, exactly, constitutes prudence?
Trustees and fiduciaries have long been legally required to make such investments and only such
investments as a prudent man would make of his own property.7 For fiduciary investors, state statutes or
federal pension law govern the approaches that are acceptable and defensible. Fiduciary law is dynamic, as
courts interpret legal requirements, and legislatures define and refine the trust investment process.
Increasingly, trust law relies on developments in the academic community to ensure that individuals
responsible for investing funds for others follow a sound decision making process.
Prior law restricted trustees investment flexibility through the imposition of legal lists of approved
investments, or through implicit endorsement of investments used by most other trustees (safety in
numbers). This approach meant that trustees could find safety from liability primarily in extremely low risk,
low return investments. Too often, however, the perverse result was that, after inflation and taxes, trust
estates depreciated in value. To correct this problem, trust law was restated in the 1990s. Many state
legislatures have enacted trust law based on new Prudent Investor Standards promulgated by the Third
Restatement of the Law (Trusts).8
For example, the California Uniform Prudent Investor Act, which became effective on January 1, 1996,
represents the states explicit endorsement of many of the concepts underlying Modern Portfolio Theory. The
investment principles embodied in the Act apply to any investor concerned with prudent wealth management:
In evaluating the prudence of any individual investment, the investment must be considered
as a component of the overall trust portfolio, rather than in isolation;
The tradeoff between risk and return should be the fiduciarys principal consideration;
Ellis, Charles D., Investment Policy. Business One Irwin Homewood, Illinois (1985), pp. 53-54.
Restatement of the Law, Second, of Trusts, 227 (1959)
8 Restatement of the Law, Third, of Trusts - Prudent Investor Rule (1992)
Uniform Prudent Investor Act National Conference of Commissioners on Uniform State Laws
Chicago October 1994. Adopted by the State of California July 5, 1995
6
7

No investment is deemed imprudent per se; consequently, the trustee may invest in any
instrument that would play an appropriate role in achieving the trusts objectives, provided
that it meets the requirements of prudent investing;
Fiduciaries must diversify the trusts investments unless it is prudent not to do so;
Trustees may delegate responsibilities for investment management to appropriately selected
qualified third parties;
The fiduciary must balance the need for current income with protection of purchasing power;
A prudently managed portfolio avoids unjustified expenses.
The updated and revised Prudent Investor Rule frees trustees from the straitjacket of low risk / low return
investments, and gives them broad latitude to invest in essentially any asset. The price they pay for this liberty
is adherence to standards of prudence that require use of appropriate care, skill and caution in the design,
implementation and management of portfolios. A prudent asset management process considers
diversification, asset allocation, risk management and cost control to be critical components of investment
success for any modern portfolio. A prudent investment management process contemplates something more
than looking for the investment or investment manager with the best track record. SCLC believes that
portfolio design and asset management deserve the same level of care, skill and caution required by prudent
fiduciaries investing money for the benefit of others.

MARKET EFFICIENCY
In a seminal work published in 1970, Eugene Fama argued that the U.S. stock market is efficient, in the
sense that the current price of every security fully reflects all available information that could have possible
bearing on its market valuation.9 Information includes all knowledge of past price movements and all publicly
available information such as that found in corporate financial statements, government and industry reports,
management announcements, etc. This information cannot help anyone develop a trading strategy (after
accounting for research and transaction costs) that generates abnormal or excess profits because the
securitys current price already reflects all known information. The rate of such incorporation has accelerated
in recent years through widespread use of computer and communications technologies.10
David Friedman11 draws a useful analogy between the investor confronting an efficient market and a
commuter deciding whether to change lanes:
When traffic gets heavy, your lane is always the slow one. You switch. A few minutes later, the
battered blue pickup that was just behind you in the lane you left is in front of you. To
understand why it is so difficult to follow a successful strategy of lane changing, consider that
other people are also looking for a faster lane and cars moving into a fast lane slow it down,
just as people moving into a short line in the supermarket lengthen it. In equilibrium, all lanes
are equally slow.
Similarly, as each datum of new information becomes known to the markets, the first few traders to obtain
it, correctly gauge its effect on prices, and execute trades accordingly. In the terms of the traffic analogy, they
gain a slight advantage in speed over other commuters by being the first few drivers to move into the faster
lane. As the datum spreads almost instantaneously through the population of traders, the advantage of
trading diminishes rapidly. It disappears altogether when prices fully reflect the new datum. By that point,
Fama, Eugene, Efficient Capital Markets: A Review of Theory and Empirical Work, Journal of Finance (May, 1970), pp.
383-417.
10 Froot, Kenneth A. and Perold, Andre F., New Trading Practices and Short-Run Market Efficiency, Working Paper #3498.
National Bureau of Economic Research. Cambridge, MA (1990).
11 Friedman, David, Hidden Order, Harper Collins Business, 1996, NY, NY.
9

however, trades based on that (no longer novel) datum are still working their way through broker/dealer back
offices, en route to the trading floor. When these tardy trades are executed, they generate relative decrements
to return, just as the last cohort of drivers to change lanes find their prior lane outpacing them.
For individual investors, the same phenomenon plays out time and time again. When a fund manager
begins to accumulate a good track record, a hoard of investors rushing to chase returns rapidly dilutes his
future success. Cash flows into the fund rise dramatically, and the manager must work harder to apply market
insights over a broader range of buys and sells. Furthermore, if the manager truly possesses skill, he or she
will raise fees to reflect the fact that skillful managers are both highly valued and in short supply. At the end of
the day the manager, rather than the investor, captures the economic rents associated with investment skill.
In short, the market-beating investor must discover the skillful manager before there is enough data to confirm
the exisence of his skill. Given that it may be harder to pick a good manager than a good stock, this is a neat
trick. There are now more mutual funds than stocks listed on the New York Stock Exhange!
The implications of Famas argument are profound. If the market prices assets efficiently and rationally,
then there are neither hidden nuggets of undervalued stocks, nor overvalued securities ripe for a fall, and
futile attempts to beat the market are a waste of the investors time and money.
Because the decision to buy just one stock is also a decision to forgo the investment opportunities of all
other stocks, the price of each security implicitly reflects the prices of all other securities, and of all available
information about them. That is, each securitys price includes information about:

its own expected return;

the expected returns available from all other securities; and,

the degree of uncertainty (risk) surrounding each securitys return forecast.

Thus each individual securitys price discounts for the unique risk/return factors of all securities.
Consequently, securities prices are net of risks unique to any specific investment, and are discounted for
differential expected returns to different securites.
In an efficient market, all assets have the same risk-adjusted returns (the discount rate for more risky
investments is higher than the discount rate for less risky investments). If a security offered a risk/reward
tradeoff more attractive than that offered by the market as a whole, profit maximizing investors would,
according to the theory, sell their positions in the market (thus lowering the market price) to buy positions in
the more attractive security (thus bidding up its price). This adjustment is made quickly and continues until the
risk/reward equilibrium among all securities is restored. In the language of economists: an efficient market
offers no projects with the expectation of positive net present value. The upshot of this line of argument is
that, on average, the current market price of any security should be close to its economically justified price. It
is difficult to beat the market simply because it is difficult to identify mispriced securities.
Although elegant and mathematically compelling, the Efficient Market Hypothesis remains controversial.
Many portfolio managers claim that they have the skill to form portfolios with expected excess profiti.e.,
returns higher than commensurate with the risks they take. However, earning a high return by assuming great
risk is not a sign of skill anymore than earning a low return by investing in conservative investments signifies a
lack of skill. The problem is that, whlile in any period there are managers who beat the market, their ranks are
not at all stable. Professional managers may be correct in thinking they can beat the market today, but this
claim is difficult to prove; and, is it not easy to identify which managers will outperform during future periods. A
prudent choice requires a careful statistical analysis encompassing much more than a nave examination of
the realized sequence of investment returns (track record). If the realized sequence of returns is the product
of luck, placing wealth in the hands of such a manager is not prudent and may lead to unpleasant
consequences.

The key question, then, is whether to try to beat the market. Many investors seek to solve intertemporal
cash flow timing problems.12 Borrowers use financial instruments like mortgages to move money from the
future to the present; savers (buyers of financial assets) move money from the present into the future. A
retirement accumulation portfolio, for example, is appropriate for individuals with current surplus labor income
wishing to support future retirement consumption from financial asset income. The purchase of a Certificate
of Deposit provides future interest earnings, the purchase of a stock provides the expectation of future
dividends, the purchase of commercial real estate provides the expectation of future lease income, and so
forth. Investors will sell the assets when they choose to bring the future income streams back into the present
(redeem the money previously moved forward in time).
A critical decision point for all investors is how they seek to solve their intertemporal cash flow problems.
If investors have no special insights or trading skills, or are unable to find profitable opportunities to exploit,
then they will tend to use investment strategies that generate market-based returns. If investors have private
information or unique skills that lead to a credible expectation that they can earn a higher-than-market return,
they will forgo purchasing diversified, market-based financial instruments in favor of establishing more
concentrated investment positions. However, generating excess returns may require great levels of skill (skills
must not be merely better than average; but require the investor to be better than other competing
professionals also seeking to beat the market).13
Thus the decision to employ an active management investment strategy14 is warrented when the investor
is confident either that he or she has a skill set sufficient to identify and profitably exploit investment
opportunities; or, has identified managers with the requisite skills. The Prudent Investor Rule for trustees
provides guidelines appropriate for individual investor portfolios:
Active strategies, however, entail investigation and analysis expenses and tend to increase
general transaction costs, including capital gains taxation. Additional risks also may result
from the difficult judgments that may be involved and from the possible acceptance of a
relatively high degree of diversifiable risk. These considerations are relevant to the trustee
initially in deciding whether, to what extent, and in what manner to undertake an active
investment strategy and then in the process of implementing any such decisions.
If the extra costs and risks of an investment program are substantial, these added costs and
risks must be justified by realistically evaluated return expectations. Accordingly, a decision
to proceed with such a program involves judgments by the trustee that: (a) gains from the
course of action in question can reasonably be expected to compensate for its additional
costs and risks; (b) the course of action to be undertaken is reasonable in terms of its
economic rationale and its role within the trust portfolio; and (c) there is a credible basis for
concluding that the trusteeor the manager of a particular activitypossesses or has access
to the competence necessary to carry out the program..15
Given the academic presumption of market efficiency, hiring a manager to beat the market is prudent only
if careful analysis documents consistent and persistent investment skill. A credible analysis, however, requires
a sophisticated set of statistical tools with which to evaluate an historical track record.

Harris, Larry, Trading & Exchanges: Market Microstructure for Practitioners (Oxford University Press, 2003), p. 178.
Ibid., pp. 475-476.
14 A passive investment management strategy consists of tracking the market without attempting to anticipate its
evolution; active management, by contrast, is the attempt to perform better than the market primarily through security
selection or market timing.
15 Restatement of the Law, Third, of Trusts - Prudent Investor Rule (1990), Comment h (Prudent investment: theories and
strategies).
12
13

10

There is a risk/return tradeoff implicit in choosing any investment management strategy. Active
investment management may provide additional funds at the cost of assuming a higher risk of failing to
achieve the portfolios required return. 16 On the other hand, passive management secures market-based
returns in broadly diversified portfolios (thus avoiding risky bets); but may limit the investors ability to earn
excess returns. The investor should develop a considered opinion about market efficiency, and should
document the asset management approach in the Investment Policy Statement. When making the choice
between passive and active investment management it is vital to determine if your investment goal can be
defined as an attempt to solve an intertemporal cash flow problem or as an attempt to beat the market.
Although the attempt to beat the market involves a speculative dimension that may or may not be justified in
terms of the underlying investment objectives, it is nevertheless inappropriate to rule out such a strategy as
imprudent. Indeed, if the returns of a benchmark index such as the S&P 500 bear little relation to the
investors personal economic objectives, then a passive investment in such an index may be every bit as
imprudent as hiring an active manager based primarily on the managers historical track record.
RISK
Securities pose two basic types of risk:
Systematic risk (sometimes called market risk), ...due to common factors facing all firms in
the economy and/or industry: the business cycle, interest rates, inflation, and so on.17; and,
Unsystematic risk (also known as unique risk), the risk unique to each firm, such as the
possibility of labor strife, litigation, product obsolescence, raw material scarcity, and
management ineptitude.
The global capital market, consisting of all the worlds capital allocated among all available investments,
and considered as a single aggregate portfolio, is devoid of unsystematic risk. It is not uniquely risky to make
any particular investment if one has made all available investments, weighing each against all the others. This
is to say, a portfolio consisting exclusively of U.S. stocks and bonds carries the economic risks of unwise
domestic fiscal and monetary policies. A global portfolio, however, mitigates such risks.
An investor must be compensated for bearing risk. But, considered as a single investment, the global
capital market has diversified away all unsystematic risk, and thus demands no compensation for it. Thus, the
global capital market sets prices without any expectation of bearing unsystematic risk. Under classic capital
market theory, individual investors paying full market prices cannot expect to be rewarded for bearing
unsystematic risk by owning only selected portions of the market. Because the future pattern of returns of a
given security is unpredictable, those returns are distributed stochastically across the whole population of
investors. Tomorrows price for, e.g., GM stock may reward an investor who owns nothing else, or those who
own none (i.e., everything but GM). The risk of owning nothing but GM, on the other hand, is borne completely
by the owner. Unsystematic risk is, therefore, ultimately defined as uncompensated risk. The more effectively
a portfolio is diversified, the less uncompensated risk it bears.
The fully diversified individual investor choosing to invest globally has an unconditional expectation of
reward as measured by the expected return (price of risk) offered by risky investments multiplied by the
amount of risk taken by investor [expected reward = (price of risk)(amount of risk)]. Investors placing money
with active managers, however, have only a conditional expectation of reward. Their expectation is
conditioned, of course, on the likelihood that they have, in fact, identified a truly skilled manager.
16 In the 2004 Chairmans letter to shareholders of Berkshire Hathaway, Warren Buffett observes: Over the last 35 years,
American business has delivered terrific results. It should therefore have been easy for investors to earn juicy returns: All
they had to do was piggyback Corporate America in a diversified, low-expense way. An index fund that they never touched
would have done the job. Instead, many investors have had experiences ranging from mediocre to disastrous.
17 White, Gerald, Sondhi, Ashwinpaul., and Fried, Dov, The Analysis And Use Of Financial Statements.
John Wiley & Sons New York (1994), p. 294.

11

Two conclusions flow from these observations, both of which are embedded in the Prudent Investor Rule:
The riskiness of any particular investment cannot be judged in isolation, but only in terms of
its effect on the portfolio (indeed, the American Law Institute cautions trustees that there are
no safe investments because even short-term U.S. Treasury securities involve certain types
of risk); and,
Failure to diversify on a reasonable basis in order to reduce uncompensated risk is ordinarily
a violation of both the duty of caution and the duties of care and skill.... Diversification is
fundamental to the management of risk and is therefore a pervasive consideration in prudent
investment management. So far as is practical, the duty to diversify ordinarily applies even
within a portion of a trust portfolio....18
Beyond the academic and legal discussions of investment risk, however, lies the more fundamental
concept of risk from the investors perspective. Without taking risk, the investor can expect to earn only the
risk-free rate of return, defined as the return on short-term default-free securiries like a FDIC-insured certificate
of deposit or a U.S. T-Bill. Unfortunately, the risk-free return is usually insufficient to keep pace with inflation
after paying taxes. In other words, the real, after-tax risk-free rate of return has historically been close to 0%.
Prudent portfolio management is a not a process of avoiding risk altogether (this would also mean avoiding
return), or of ignoring risk (this is required for strategies designed to maximize return). Prudence requires that
risk be measured and managed.
Unfortunately, such measurement and management remains an especially difficult task. Most investors
conceptualize risk in terms of fuzzy labels: low risk tolerance, safe, average risk tolerance, moderate,
aggressive. These Ambiguous characterizations, however, must be converted operationally to explicit and
readily understandable quantitative measures. Often, risk measures may be expressed in terms of volatility:
the annual standard deviation of the portfolio must not be greater than x; in terms of probabilities: the
probability of a loss over a specified period should not exceed y%; in dollar terms: the likelihood of a decine in
value equal or greater than $$$ should not be more than z%; or, in terms of failure rates or confidence
intervals: the portfolio has an x% confidence interval with respect to achieving this economic goal. Other
quantitative measures (range, shortfall probability, tracking error v. comparable benchmark, etc.) may also
have useful applications. 19 No matter how you express the concept of investment risk, however, the
fundamental point remains: if portfolio managers are not managing portfolio risk, they are not managing
portfolios.20

DIVERSIFICATION
Harry Markowitz, investigating the question of how to invest under conditions of uncertainty, completed
the seminal work on portfolio diversification in the early 1950s. There is more to diversification, he noted,
than simply buying many different investments. Intelligent diversification takes into account the contribution
of each investment to the risk and return characteristics of the entire portfolio. According to Markowitz, to
assess that contribution, the investor must consider the following factors:
The expected return of the asset. This refers to the statistically most probable (mean
expected) rate of return. The return expectation can be a forecasted return, it may be based
on the assets return history;
Restatement of the Law, Third, of Trusts (Prudent Investor Rule), Chapter 7 pp. 18, 23 and 25.
Investors may be faced with many risk variables including inflation risk, longevity risk, tax and regulatory risk, labor
income interruption/termination risk, etc.
20 Sykes Wilford, D., Risk Measurement versus Risk Management, Improving the Investment Process through Risk
Management, (Association for Investment Management and Research, 2003), p. 17.
18
19

12

The standard deviation of the asset. This refers to either the forecasted or the historical
variability of the assets returns. The higher the standard deviation, the greater the probable
variance of any one periods return from the expected return. For example, an asset with an
expected return of 10% and a standard deviation of 12 would be expected to generate actual
returns of between -2% and +22% about 68% of the time, and between -14% and +34%
about 95% of the time [ two standard deviations];
The correlation of the assets returns to returns of other assets within the portfolio. The
correlation coefficient measures how similarly the returns of any two assets behave under
similar economic conditions. A correlation coefficient of perfect unity indicates that they will
move in tandem, and that no risk reduction benefits can be achieved by their combination.
Conversely, A correlation coefficient close to zero, or negative, indicates that the return
patterns are not closely linked, and thus that a combination of such investments may
significantly reduce overall portfolio risk. The value of the correlation statistic is an average
taken over many periods and, for any single period, the actual co-movement of asset returns
may differ from the average.
Figure 1.1 depicts returns from three investments. A and B exhibit perfect negative correlation: Bs pattern
of returns is the exact inverse of As. A and C, on the other hand, exhibit positive correlation: they both tend to
move in the same direction, to varying degrees, under similar economic circumstances.21

The long-term historical correlation of several asset classes may be seen in Figure 1.2.

21 Technically, negative correlation indicates that when asset As return is above its mean, asset Bs return tends to be
below its mean. Thus, in the same period, both investments might exhibit positive returns but remain negatively correlated.
Negative correlation does not mean that when asset A increases in value, B decreases in value. Such a portfolio would go
nowhere fast!

13

Fig ure 1.2: Correlation Matrix for Returns from Investment Categ ories
1973 through 2006

S&P 500 Stocks

US Large Value
Stocks*
US Small Stocks**
US Small Value
Stocks***
MSCI EAFE Int'l
Stocks
Int'l Small
Stocks****
MSCI Emerging
Markets
NAREIT Real
Estate Index
Citigroup World
Bond Index

# of
US Large
US Small
MSCI
MSCI
Months S&P 500
Value
US Small
Value EAFE Int'l Int'l Small Emerging
Data
Stocks
Stocks
Stocks
Stocks
Stocks
Stokcs
Markets
397
1.00
0.85
0.81
0.75
0.55
0.40
0.59

NAREIT Citigroup
Real
World
Estate
Bond
Incex
Index
0.52
-0.03

397

0.85

1.00

0.76

0.84

0.48

0.37

0.51

0.59

-0.08

397

0.81

0.76

1.00

0.92

0.50

0.45

0.65

0.63

-0.08

397

0.75

0.84

0.92

1.00

0.48

0.44

0.61

0.66

-0.11

397

0.55

0.48

0.50

0.48

1.00

0.85

0.58

0.36

0.43

397

0.40

0.37

0.45

0.44

0.85

1.00

0.54

0.31

0.40

228

0.59

0.51

0.65

0.61

0.58

0.54

1.00

0.32

0.00

397

0.52

0.59

0.63

0.66

0.36

0.31

0.32

1.00

0.02

264

-0.03

-0.08

-0.08

-0.11

0.43

0.40

0.00

0.02

1.00

* Fama French
Large Value
** CRSP 6-8
Decile
*** Fama French Small Value
**** DFA International Small Index

When investments are evaluated in isolation, their risk and return characteristics provide little information
about the effect they will have on a given portfolio.22 The correlation structure of the securities to be included
in the portfolio is a key determinant of long-term investment successyet, it often remains invisible to the
average investor; or ignored in the chase to catch high returns.
Comprehensive diversification makes the measurement and control of overall portfolio risk possible. A
portfolio concentrated in a few securities is exposed to so much unsystematic risk that it is difficult to reach
reliable judgments regarding the range of expected future returns. Consider the example of emerging markets
during the period 1988 through 2006. As maybe seen in Figure 1.3, emerging markets were extraordinarily
volatile over the period: 23

22
23

Gibson, Roger C., Asset Allocation: Balancing Financial Risk (Business One Irwin, 1990), p. 118.
International Finance Corporation Data reported by Ibbotson Associates (Chicago, Ill.).

14

Figure 1.3: Emerging Country Returns

1988 through 2006

1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
1988-2006

Brazil
129.9%
81.1%
-68.4%
161.8%
9.6%
73.0%
63.4%
-19.3%
42.5%
27.4%
-39.3%
66.2%
-11.3%
-17.0%
-30.6%
114.5%
36.8%
56.9%
45.9%
23.4%

Korea
96.8%
1.4%
-27.1%
-15.5%
1.1%
31.5%
23.8%
-3.1%
-38.2%
-66.6%
140.6%
92.0%
-49.5%
48.7%
8.6%
35.8%
23.0%
58.5%
12.8%
8.3%

Mexico Indonesia Thailand Portfolio

66.3%
258.1%
47.5%
119.7%
91.7%
81.9%
114.2%
74.1%
63.1%
5.9%
-27.4%
-10.8%
125.2%
-44.0%
23.9%
50.3%
25.6%
-0.6%
34.3%
14.0%
49.2%
106.1%
105.0%
73.0%
-39.9%
-26.2%
-9.2%
2.4%
-21.3%
8.9%
-3.8%
-7.7%
18.6%
27.9%
-36.8%
2.8%
54.0%
-67.4%
-73.4%
-25.2%
-33.6%
-44.8%
10.9%
6.8%
80.4%
92.7%
47.7%
75.8%
-20.9%
-62.2%
-56.2%
-40.0%
19.4%
-8.5%
5.3%
9.6%
-12.8%
42.8%
27.5%
7.1%
32.0%
78.2%
144.7%
81.1%
48.0%
52.2%
-1.0%
31.8%
49.0%
15.9%
9.2%
37.9%
41.7%
74.7%
13.9%
37.8%
25.7%
11.3%
6.6%
22.0%

The average annualized return for these eight countries was 15%. On the other hand, an investor who
followed a strategy of investing an equal percentage of a portfolio in each of these markets, rebalancing at the
end of each year to restore that initial allocation, would have experienced a compounded annual return of
22%. The return premium due to the diversification (asset allocation policy) and risk control (rebalancing
policy) was, therefore, 7%.
The effect of the diversification strategy may be seen in Figure 1.4:

15

Fig ure 1.4: Growth of $1,000 in Emerging Markets

1 988 through 2 006

Mexic o
$77,249

$80,000
$70,000

Brazil
$54,058

$60,000
$50,000

Por tfo lio

$ 43,993

$40,000

A verag e
$ 28,375

$30,000
$20,000
$10,000

Ind onesi a
$7,643 Korea
$4,554

$1,000

Thaila nd
$3,342

1988

1990

1992

1994

1996

1998

2000

2002

2004

2006

The risk posed by a diversified portfolio is less than the weighted average risk of its individual assets. An
effectively diversified portfolio contains assets that respond differently to new information. Their patterns of
returns tend to offset or neutralize each other. The expected return to a portfolio, on the other hand, is the
weighted average of the returns expected for each of its indiviudal assets.24 The reduction in risk due to
diversification does not result in a reduction in return. Effective portfolio diversification thus provides investors
with rewards commensurate with a level of risk higher than they actually bear: This is a major reason why
diversification is valued and why the prudence of a trustees investment is to be judged by its role in the trust
portfolio rather than in isolation.25
Comprehensive diversification makes possible the measurement and control of overall portfolio risk. A
portfolio concentrated in a few securities is exposed to so much unsystematic risk that reaching reliable
judgments regarding the range of expected future returns is difficult. In Figure 1.4, the Indonesian market was
down 44% in 1991, when the global trend in emerging markets generally was up (+50.3%). Indonesia in 1991
was not representative of emerging markets. Likewise, for any grouping of securities that are somewhat
similar, and for any period, none of them are necessarily strongly correlated with the others, or with the group
as a whole. By examining the behavior of any one of them, the investor can discern little about the behavior of
all of them put together.
When, on the other hand, a portfolio owns enough securities to allow for statistically significant sampling,
the investor can reach reliable judgments about risk and return. The amalgamation of the whole array of
emerging markets results in what is effectively a virtual security, with its own risk and return characteristics,
different from those of any of its components. The same is true for comprehensively diversified holdings of any
other type of security, such as domestic small company stock or municipal bonds. This is why investors are
interested in the published indices (e.g., the S&P 500, the Lehman Brothers Corporate & Government Bond
Index, etc.) which reflect comprehensive diversification across a class of securities, or asset class. The
behavior of these indices provides a point of reference, or benchmark, for the performance of securities of the
same type.

24
25

Farrell, James L., Portfolio Management: Theory and Application (Irwin McGraw-Hill, 1997), p. 22.
Restatement of the Law, Third, of Trusts (Prudent Investor Rule), The American Law Institute (1992), p. 26.

16

A virtual security is likewise created when diversified holdings across different asset classes are
amalgamated into a portfolio. The risk and return characteristics of the portfolio, derived from those of its
component asset classes, differ therefrom. However, it is difficult to calculate these portfolio characteristics.
In practice, such calculations require specialized software.26 Unless the investor performs such calculations,
however, the overall character of the portfolio, and particularly the risk it poses, remains hidden. Yet these
hidden data are the most important insights an investor can obtain. Without knowledge of overall portfolio risk
and return, the investor is effectively blind with respect to the consequences of any particular investment
decision.

ASSET ALLOCATION
The American Law Institutes explanation of the Prudent Investor Rule points out that, asset allocation
decisions are a fundamental aspect of an investment strategy and a starting point in formulating a plan of
diversification.27 The portfolios asset allocation describes its percentage distribution among different kinds of
investments or asset classes.28 Asset allocation may be an important determinate of portfolios expected longterm rate of return and the degree of risk assumed to achieve it.
The importance of the asset allocation decision on long-term investment results is a subject of some
current controversy. Conventional wisdom, based on a 1991 study, suggests that the asset allocation decision
is the primary determinant of return variance for portfolios with long-term planning horizons, as is shown in
Figure 1.5.

26 For independent trials, the variance of a sum (i.e., a portfolio) equals the sum of the variances (i.e., the individual
components). However, because of the correlation structure of securities within the portfolio, the variance of a sum equals
both the sum of the variance and the sum of all cross-product (or covariance) terms. In a two asset portfolio, there are
two individual asset variance terms plus two covariance terms; in a four asset portfolio there are four individual asset
variance terms plus twelve covariance terms; in a ten asset portfolio there are ten individual asset variance terms plus
ninety covariance terms; in a hundred asset portfolio there are one hundred individual asset variance terms plus nine
hundred and ninety covariance terms. In the limit, as the number of assets grows large, the proportional risk of any
individual asset moves asymptotically towards zeroonly systematic or market risk remains. Labels like safe assets or
growth assets become meaningless because investments cannot be evaluated in isolation but only from within the
portfolio context.
27Restatement

(Third) of The Law Trusts: Prudent Investor Rule The American Law Institute (1992)., p.25.

One commentary suggests that each asset class should include relatively homogeneous investments, and the asset
classes should be mutually exclusive. For the purposes of risk control, an included asset class should not have extremely
high expected correlations with other asset classes (or combinations of other asset classes). From a similar perspective, it
is also desirable that the asset classes together make up a preponderance of world investable wealth. The authors point
out the use of asset classes within the strategic asset allocation context: Distinct (and well-differentiated) groups of assets
should have distinct exposures to factors and/or exposures to different factors. These observations suggest a key
economic role of strategic asset allocation: A strategic asset allocation specifies the investors desired exposures to
systematic risk. Sharpe, William F., Chen, Peng, Pinto, Jerald E. & McLeavey, Dennis W., Asset Allocation, Managing
Investment Portfolios, (John Wiley & Sons, 2007), pp. 230-320.
28

17

Figure 1.5

Fi gu re 1.5: D ete rmi n an t s of P or tfo li o P erf orm an ce

93.60%

As s et Alloc ation

Security Selection

Market Timing

Other F ac tors

4.20%

1.70%

0.50%

From a short-term perspective, these findings are counterintuitive, since over the short run, stock selection
and transaction timing have a significant impact on returns. But focusing on the short term can be detrimental
for investors with longer planning horizons. In fact, the study cited above found that market-timing activities
actually subtracted returns from portfolios over planning horizons longer than ten years. The study, which
examines large pension plans, concludes that asset allocation explains 90% of the variation in return over time
of the average pension plan.29 However, the study is controversial because an averaging function is also ipso
facto a smoothing function. Therefore the study does not focus on the time series of returns for individual
retirement plan portfolios, nor does it explain the reasons why individual portfolio returns differ from each
other.
By contrast, some recent theoretical studies suggest that, even over long planning horizons, security
selection should dominate asset allocation decisions with respect to its affect on portfolio performance. For
example, a 2003 study of international investment decision making outlined five factors that could explain
investment returns: (1) asset allocation, (2) country allocation, (3) global industry sector allocation, (4) countryspecific industry sector allocation, and (5) security selection. The authors isolated each factor and simulated
10,000 portfolios (60% stock/40% fixed income asset allocation) using data based solely on the single factor
from 1987 through 2001. Portfolios based on the security selection factor had the greatest range of returns;
portfolios based on the asset allocation factor had the smallest dispersion of returns. Therefore, the authors
conclude that security selection has the greatest potential for influencing long-term investment returns.30

Brinson, Gary P., Singer, Brian D., and Beebower, Gilbert L., Determinants of Portfolio Performance, Financial Analysts
Journal May-June, 1991)
30 Kritzman, Mark & Page, Sebastien, The Hierarchy of Investment Choice, Journal of Portfolio Management (Summer,
2003), pp. 11-23.
29

18

Other recent studies come to a different conclusion. For example, a 2000 study of the affect of asset
allocation on investment performance points out that the importance of asset allocation depends on the
investment issue under consideration.31 Specifically, the investor might be interested in knowing:
1. What percentage of a portfolios ups and downs (variability in return) is explained, over time, by its
asset allocation choices? or,
2. How much of the performance difference between two distinct portfolios can, over time, be
explained by differences in their asset allocation? or,
3. How much of a specific portfolios actual returns can, over time, be explained by its asset
allocation?
These are very different questions and require separate methods of analysis. The authors decompose the
monthly returns of balanced mutual funds over a ten-year period into a policy return (the return attributable to
the funds asset allocation, and an active return (the remaining return or: {total actual return policy return}).
The study confirms that approximately 90% of the variability in the returns of the average (median) fund can be
explained by its asset allocation decisions. When funds are compared to each other, however, the conclusions
differ. If two funds select the same asset allocation and each invests in the same cross-section of passively
managed indexes, 100% of the variability of returns across time of each fund would be attributable to asset
allocation policy.32 In fact, however, the mutual funds under evaluation differed with respect to both their
asset allocations and their security selection, market timing, fees and other factors. The study concludes that,
on average, asset allocation decisions account for about 40% of the variation of returns across funds.
Finally, the authors test for the percentage of individual fund returns33 that, over time, can be explained by
asset allocation. This is the ratio of policy return divided by total actual return. A hypothetical fund with a
consistent asset allocation policy implemented by a purely passively investment strategy will, by definition have
a ratio equal to one. Funds exhibiting ratios greater than one will have subtracted value through active
management decisions (actual total returns in the demoninator fail to equal the policy returns in the
numerator); funds exhibiting ratios less than one will have added value through market timing (decisions to
change asset allocation to exploit forecasted market developments) or security selection. The distribution of
ratio values is very interesing. The median result (50th percentile) was 1.00on average, actively managed
mutual funds neither added nor subtracted value during the period under evaluation.34 The best actively
managed funds (5th percentile) exhibit ratios of 0.82; however, the worst performing funds exhibit ratios of
1.32. But greater dispersion of results is simply another expression for investment risk and uncertainty.
The authors derive two conclusions from these results:
1. Because the active managers, as a group, cannot achieve a return greater than the return of the
market (the average performance before costs of all investors must equal the performance of the
market), asset allocation policy explains, on average, approximately 100 percent of the returns of
mutual funds.

31 Ibbotson, Roger G., & Kaplan, Paul D., Does Asset Allocation Policy Explain 40, 90, ore 100 Percent of Performance?
Financial Analysis Journal (January/February, 2000), pp. 26-33.
32 Likewise, if two funds had the same asset allocation policy but each invested in a separate set of securities, asset
allocation would explain 0% of the return differences over time.
33 As opposed to the percentage of the variability of return.
34 Results are pre-tax. For taxable investors actively managed funds may trigger substantial income tax liabilities because
of their higher level of turnover. It is interesting to note that the inability to add value is also a test of the efficient market
hypothesis. In this case, markets are considered efficient if the profits derived from active management are unable to
overcome extra costs and risks. Forecasts generate, on average, zero net profit. The distribution in this study is skewed
towards the downside indicating that finding superior investment managers is a difficult task.

19

2. If the investor has the ability to select superior managers before committing funds, there is a
possibility of earning market-beating returns. This entails not only a close examination of riskadjusted historical results, but also the assumption that such results will persist into the future.
The implications for investment policy are clear:
Over longer planning periods, the asset allocation decision is an important factor in
determining returns;
The choice of active management may be prudent; however, the investor should be aware
that the attempt to beat the market is, itself, a significant contributor to portfolio risk; and
Long-term policy should be designed to insulate the portfolio, cushioning the impact of
business and market cycles, and forestalling ill-considered decisions based on short-term
factors. Abandoning policy increases portfolio risk by subjecting assets to the vagaries of
transitory economic conditions.

20

C H A P T E R T W O : A S S E T C L A S S E S A N D A S S E T C L A S S I N V E ST I N G

In the treasure hunting model, the money manager may concentrate on purchase and sale of indivdual
securities (stocks & bonds). By contast, the prudence model emphasizes asset class investing rather than
security selection and timing of buy/sell transactions.35 Asset classes may represent the broad cross section
of all (or most) securities within a capital market. A commonly used vehicle for such investments is index
funds, although there are many investment product variations on this theme.
An asset class is a building block of a portfolio. An asset class:

Consists of securities that exhibit common statistical, economic or accounting characteristics.

Is expected to exhibit differing risk/reward responses to changes in economic conditions.

May be generally categorized into two macro groupings; fixed income (bonds), and equity
(stocks).

U.S. EQUITIES
Historically, returns from U.S. stocks have far outpaced those from investment grade U.S. corporate and
government debt. For example, one dollar invested in common stocks (as represented by the Standard &
Poors 500 stock index) at the beginning of 1926 would have been worth $2,533.20 (assuming dividend
reinvestment) by the end of 2004. The same dollar invested in long term U.S. Government Bonds would have
been worth just $65.70. If invested in U.S. Treasury Bills (30-day), the dollar would have grown to only $17.90.
Inflation over this period required an increase to $10.60 simply to maintain purchasing power. The greatest
return on investment over this period was produced by small stocks36 which increased the value of each 1926
dollar to $13,641.20 by the end of 2004.37

Under certain conditions, it may be prudent to eschew broad-scope diversification. This is, however, a complex issue \
beyond the scope of an introductory essay on portfolio management. Interested readers may find further information in the
working paper: Prudence which is currently posted on the firms website at schultzcollins.com.
36 As measured by the 9th and 10th deciles of the Center for Research in Securities Prices [CRSP] database.
37 Ibbotson Associates, (Chicago, 2005).
35

21

The fact that equity investments have outperformed fixed income investments over long time horizons is
demonstrated empirically in the following table:

Fi g ur e 2 .1 : No m in a l A nn u a l ize d Re tu r n s
1926 through 2006
T ime Period # of Years
1926-2006
81
1957-2006
50
1967-2006
40
1977-2006
30
1987-2006
20
1997-2006
10
2002-2006
5

S&P 500
10.42%
10.62%
10.98%
12.47%
11.80%
8.42%
6.19%

U .S. Long Term Government Bonds Inflation (C PI)

5.42%
3.03%
6.77%
4.06%
7.77%
4.64%
8.97%
4.23%
8.61%
3.06%
7.83%
2.44%
7.19%
2.69%

The historical advantgage of equities is even better domonstrated when returns are adjusted for inflation.

Fig u r e 2 .2 : In fla ti o n A d ju ste d A n n u a lize d R e tu rn s

1926 t hrough 2006
T ime Period # of Years
1926-2006
81
1957-2006
50
1967-2006
40
1977-2006
30
1987-2006
20
1997-2006
10
2002-2007
5

S&P 500
7. 17%
6. 31%
6. 06%
7. 90%
8. 48%
5. 84%
3. 41%

U .S. Long Term G overnment Bonds

2.31%
2.61%
3.00%
4.55%
5.39%
5.27%
4.38%

I nflation (C PI)
3.03%
4.06%
4.64%
4.23%
3.06%
2.44%
2.69%

Equities have consistently and significantly outperformed inflation over extended holding periods.38
38 The reason for the magnitude of U.S. stocks outperformance of U.S. bonds, however, remains controversial. Economic
theory suggests that historical excess reward (stock return risk free rate = equity risk premium) received by owners of
U.S. stocks is abnormally high when adjusted for risk; or, abnormally high when compared to the rewards to holding stock
achieved by investors in foreign capital markets. Economists refer to this controversy as the Equity Risk Premium Puzzle.
The reader is cautioned that the empirical fact of equity outperformance is by no means a guarantee that it will continue to
outperform fixed income in the future. Mark Rubinstein, for example, asks: how long must an investor be prepared to wait
before the probability becomes high that an all-stock portfolio will outperform an all-bond portfolio? Rubinstein develops
the following theorem: Assume that all available assets collectively follow a stationary random walk in continuous time (with
finite variance). Let X and Y be the values after elapsed time t > 0 from following two strategies (with equal initial total
investment), each being the result of continuously rebalancing a portfolio to maintain constant proportions in the available
assets. Then:

Probability (X > Y) = N {

( X Y ) t
[ 2 X Y Y2 ]1 / 2
2
X

where N is a joint standard lognormal probability distribution, Xt is the expected value of log (X), Yt is the expected value
of log (Y),

X t

is the standard deviation of log (x),

Y t is the standard deviation of log (Y), and is the correlation

between log (X) and log (Y).

Assuming, based on a reasonable sample of historical data, that stocks offer a 2.5% return premium over bonds, with the
standard deviation of stocks equal to 18% and the standard deviation of bonds equal to 10% with a correlation of 0.4, in
order to be 95% confident that an all stock portfolio will outperform an all bond portfolio requires a planning horizon of 123
years. Rubinstein, Mark Continuously Rebalanced Investment Strategies, Journal of Portfolio Management (Fall, 1991).

22

Volatility and Return

Given a long-term planning horizon, equity returns are generally higher than fixed income returns. In the
short run, however, equity returns are more volatile. Equity investors must be compensated for bearing this
increased risk. The following graph highlights the ranges of annual returns for five major asset classes over the
81-year period from 1926 through 2006.39 All returns are nominal that is, they have not been adjusted for
inflation.

Figure 2.3: Best and Worst Annual Returns

1926 through 2006

142.9%

Best Year
Worst Year

54.0%
42.6%

40.4%
14.7%
0.0%
-8.1%

-9.2%

-43.3%
-58.0%
Small Stocks

S&P 500 Stocks

Corporate Bonds

U.S. Long Term

Government Bonds

U.S. 30 Day T-Bill

The widest range of returns (i.e., the most volatile asset classes) both come from equities, with small
company stock returns hitting a high of +142.9% in 1933 and a low of 58.0% in 1937. The S&P 500 has a
narrower range of historical returns. Its best month was also in 1933, when it advanced by +54.0%, while its
return was 43.3% in 1931.
Generally, fixed-income investments are less volatile. Returns from corporate bonds ranged from +42.6%
(1982) to 8.01% (1969); Government Bonds have experienced returns between +40.4% (1982) and -9.2%
(1967). U.S. Treasury Bills have never had a nominal loss greater than a small fraction of 1%, although they
have often lagged inflation. Although Treasury Bills sometimes provide good short-term inflation protection,
their long-term inflation adjusted track record is poor. Their best ten-year period was 1981 through 1990,
during which time they produced an annualized return of 3.3% above inflation. Over the sixty years between
1934 and 1993, their average annualized return lagged 0.14% below inflation.40

39
40

Ibbotson Associates, Op. Cit.

Ibbotson Associates, Op. Cit.

23

Therefore, when a portfolio with a long-term planning horizon requires returns above inflation, historical
data indicate that a significant portion of assets should be invested in U.S. equities.41

U.S. FIXED INCOME

In general, fixed income returns tend to be less variable than equity returns, and therefore, have lower
expected long-term returns. Much of the return variability in bonds and other fixed income investments is
attributable to maturity risk. The greater the period to maturity, the greater the risk to the investor. As interest
rates and issuer credit ratings change over time, the market value (that is, the net present value) of maturity
proceeds, plus interim coupon payments, fluctuates. The fundamental bond pricing theorm states that
increases in interest rates or reductions in issuer credit rating cause the market value of the bond to decrease,
and vice versa. The magnitude of the price change is directly related to the time remaining until maturity, and
may be calculated using technical measures of price sensitivity known as duration and convexity.42 Prices
fluctuate for all types of bonds, irrespective of whether they are issued by a corporation or are backed by the
U.S. Government. Indeed, when the Federal Reserve Bank raised interest rates to combat inflation during the
1980s, long-term U.S. Treasuries suffered a substanital decrease in value.43
Although bonds may exhibit considerable price variability, returns from U.S. long-term fixed income assets
have generally not matched returns from U.S. equity. This fact suggests that an optimal combination of short
term fixed income assets and equities will yield a return superior to a long-term bond portfolio. For example,
during the period 1973 through 2006 the increased returns from holding longer maturity bonds was not
compensated in proportion to their increased risk. The data suggest that the optimal maturity weighting for a
fixed income portfolio is five years or less:44

41 The reader should keep in mind that long-term averages are calculated for paper indexes lacking fees, trading costs and
other expenses. Additionally, results apply only the in absence of interim cash flows. Design and management of portfolios
subject to cash flows (e.g., retirement income distributions) is very different from design and management of portfolios
seeking to generate a high amount of terminal wealth. The former are consumption-oriented portfolios, the latter are
wealth-accumulation oriented. Thus, for example, the asset allocation decision to load for equities under a distributional
regime may be counterproductive if high inflation increases the need for high distributions during a time when inflationary
pressures decrease stock prices. Although long-term equity returns have, on average, outpaced inflation, the investor must
live with actual results not average results. This introductory essay, in the main, does not discuss portfolio design or asset
allocation decisions in the face of liabilities.
42 Duration and convexity are the first and second derivatives, respectively, of the ratio of price change to yield change.
43 Jones, Charles P. Investments: Analysis And Management. John Wiley & Sons, New York (1994), pp. 182-186.
44 This discussion assumes that the investor wishes to focus on the objective of enhancing reward per unit of risk over the
applicable planning horizon. Investors faced with consumption liabilities, however, may wish to include long-term bonds
not for their reward-to-risk characteristics but rather for their ability to hedge against future economic states that might be
detrimental to a smooth consumption path. In cases where the asset management objective requires the investor to
hedge long-term liabilities, the safe asset may be a long-term bond; the risky asset may be a short term bond.

24

Figure 2.4: Fixed Income Maturity and the Risk/Reward Tradeoff

1973 through 2006

Return

11.81%

Standard Deviation
8.75%
7.98%
6.95%

6.77%

6.09%

2.98%

U.S. 30 Day T-Bill

3.34%

One-Year Treasury
Constant Maturity

U.S. Intermediate Term

Government Bonds

U.S. Long Term

Government Bonds

Additional academic data strongly support the position that analysts are unable to forecast interest rate
changes consistently. Lengthening bond maturities to take advantage of forecasted interest rate declines,
therefore, will yield uncertain results at the cost of certain transaction expenses. Comprehensive analysis of
bond price changes relative to forecasted predictions indicates that, like equities, bonds fully reflect all
available information. This analysis also indicates that it is hard to be able to consistently forecast interest
rates with greater accuracy than a no-change model.45

Why Own Any Bonds?

In reviewing historical returns from bonds and equities, an investor may conclude that the lower expected
returns of bonds should disqualify them as an asset class, and that the portfolio should consist solely of
higher-return equities. Often, an investors stated primary objective is to maximize investment return. The
primary purpose of a multi-asset class portfolio, however, is not to maximize return (leveraged derivatives are
far more effective for generating extremely high returns), nor to eliminate risk. Rather, multiple asset class
portfolios are intended to enhance the ratio of reward to risk, so that the investor has the expectation (not the
guarantee) of adequate reward for each unit of risk assumed. As R. Charles Tschampion, Director of the GM
pension plan, writes:

Sharpe, William, and Alexander, Gordon, Investments Prentice Hall, Englewood Cliffs, New Jersey (1990), p 380. For a
recent discussion of the difficulties of accurate macroeconomic forecasting, see Cohen, Abby Joseph, Aristotle on
Investment Decision Making, Financial Analysts Journal (July/August, 2005).
45

25

Two maxims are important in portfolio management. The first is: Investment management is
not an art, and it is not a science; it is an engineering endeavor. The second is: Risk can be
managed but return cannot.46
It is safe to say that most investors do not think about enhancing ratios when they invest. They think about
increasing returns. Although many investors want to maximize returns, most would find the necessary steps to
be intolerably risky. Thus the primary reasons for including U.S. fixed income investments in the portfolio are
to:
Control portfolio risk by achieving returns that approximate the inflation rate, with relatively
low levels of variance; and,
Provide portfolio diversification by generating a different pattern of return from that of
equities.47
INTERNATIONAL EQUITY
If stock market returns in every country moved in perfect lock step, and generated equal results, there
would be no advantage to owning foreign stocks. However, the lack of perfect correlation between returns of
different foreign stock markets indicates that there may be potential advantages to including foreign stocks in
a portfolio. During the 1980s and early 1990s, studies of international stock diversification (from the
perspective of an American investor) were almost unanimous in their recommendation to hold foreign equity
within the portfolio. The primary reasons for the recommendation to diversify internationally were:
1. High returns on foreign stocks; and,
2. Low correlation between US and foreign stock market returns.
One study, for example, highlighted the less than perfect correlation between stock markets of developed
nations to the U.S. stock market (perfect correlation = 1.00) for the period February 1986 through March
1991:48

Tschampion, R. C., Financial Risk Management in Global Portfolios, Investing Worldwide VI (AIMR, 1995), p. 67.
Nobel Prize winner Paul Samuelson neatly summarizes a second line of argument against defaulting to a 100% equity
position: factual happenstance is not arithmetical necessity. Samuelson, Paul A., The Long-Term Case for Equities and
how it can be oversold, The Journal of Portfolio Management (Fall, 1994), pp. 15-24. Briefly, Samuelson argues that if
stocks must always beat bonds in the long run, then the investor must believe:
1. Long-term bonds will disappear. This result would, however, violate capital market pricing theory, which
advises holding the entire range of assets according to their market weightings because, in equilibrium,
expected returns (prices) are set so that each asset in the market clears. Because bonds are a part of the
capital market, on a risk-adjusted basis, prices should bring demand for and supply of this asset class into
alignment with all other competing assets.
2. No price/earnings ratio will ever be too high, so that equities in the future could never be said to have
expected returns that are lower than bonds. But even if historical financial data has been generated by a
stationary or stable probability process, it still does not follow that random draws from the process will create
such a preponderance of superior future outcomes that you would always opt for a 100% equity position. In
Samuelsons words: The urn from which future draws gets taken is altered endogenously by one-way
movements of market levels.
48 Divecha, A. B., Drach, J., & Stefek, D., Emerging Markets: A Quantitative Perspective, The Journal of Portfolio
Management (Fall, 1992), p. 48.
46
47

26

Figure 2.5: Correlatlion of U.S. and Foreign Equity Returns: 2/86 3/91

Country

Correlation to
U.S. Stocks

Correlation to
U.S. Stocks

Australia

0.49

Japan

0.25

Austria

0.14

Netherlands

0.69

Canada

0.82

Norway

0.55

Denmark

0.30

New Zealand

0.39

Finland

0.42

South Africa

0.24

France

0.48

Spain

0.49

Germany

0.39

Sweden

0.46

Ireland

0.47

Switzerland

0.61

Italy

0.34

United Kingdom

0.67

Country

Investing in foreign stocks enabled a portfolio to realize equity returns while significantly lowering risk. In
terms of the allocation between U.S. stocks (proxied by the S&P 500) and international stocks (proxied by the
Morgan Stanley EAFE--Europe, Australia, Far East Markets--Index) shown in the following graph, a 70% S&P
500 / 30% EAFE blend provided the optimal risk/reward combination over the period 1973 through 2006:

Fig 2.6: Risk and Return of Internationally Diversified Portfolios

S&P 500 and MSCI EAFE, 1973-2006

Ave rage Return

13.2%
100% MSCI EAFE

13.0%

12.8%
70% S&P 500/
30% MSCI EAFE
12.6%
100% S&P 500

12.4%
16%

17%

18%

19%

20%
Ris k

27

21%

22%

23%

Further studies, however, cautioned that the correlation statistic is not constant; and, that correlation
often increases during periods of worldwide market volatility.49 Correlation between return series of various
nations is a function of the timing of national business cycles. In normal periods, national business cycles are
not well synchronized and, therefore, international diversification contributes greatly to the portfolio. In times
of global economic stress, however, correlation may become more positive and the risk-reduction benefit may
diminish.
Despite the fact that early reseach papers acknowledge that global equity diversification is not a fail-safe
risk control mechanism, they provide at least two other reasons for including international equity:
1. Foreign goods represent a significant portion of U.S. citizens consumption basket. Therefore
exposure to foreign investments helps protect the portfolios purchasing power; and,
2. U.S. inflation is driven by domestic political decisions as well as by economic forces. Exposure to
foreign assets acts as a hedge against unwise U.S. fiscal policy.50
By the mid 1990s, however, the combination of lower foreign stock returns, increasing globalization of
international trade and business, and increasing correlation of domestic and international returns, caused
many to question whether holding international equity could provide the expectation of significant future
diversification benefits.51 Although the average value of the correlation statistic between the S&P 500 and
the EAFE index (Index of large company foreign stocks) is approximately 55% over the period 1973 through
2006, beginning in the mid 90s the correlation statistic moved sharply upwards (i.e. became more positive).
Although the following chart illustrates the recent increase in correlation, it also indicates that the correlation
statistic exhibits a wide range of historical values.52 Although the trend line imposed upon the data has an
upwards slope, there is no assurance that the current value of the correlation statistic will continue to remain
high.

Erb, C. B., Harvey, C. R., & Viskanta, T. E., Forecasting International Equity Correlations, Financial Analysts Journal
(December, 1994), pp. 32-45.
50 Odier, P., & Solnik, B., Lessons for International Asset Allocation, Financial Analysts Journal (April, 1993), pp. 63-76.
51 See, for example, Sinquefield, R., Where are the Gains from International Diversification? Financial Analysts Journal
(1996), pp. 8-14.
52 In 1948 the correlation between an index of foreign stocks (created by Global Financial Data using EAFE index
construction principles without rebalancing) and the US market as represented by the CRSP 1-10 index of the total US
stock market was 0.30.
49

28

Fi g 2. 7: T hree Year R oll in g Av er ag e C orrelation

S &P 5 0 0 / M SC I EA FE, 1 97 1 th roug h 2 0 0 6

1. 0
0. 9
0. 8
0. 7
0. 6
0. 5
0. 4
0. 3
0. 2
0. 1
0. 0
1973

19 76

197 9

1 982

19 85

1988

1 991

199 4

1997

20 00

200 3

2006

Projections of the future value of the correlation statistic are controversial. Some argue that it should
remain higher than its historical average because of Europes economic integration and the reduction in
currency fluctuations brought about by the introduction of a common monetary unit (the Euro). Others point to
increasing globalization or the increasing importance of sector as opposed to country influence on stock
returns.53 Others point out that the strongly positive recent correlation values are a result of events unlikely to
repeat in the future. These include the collapse of the Russian ruble, the Asian banking crisis and the
worldwide collapse of stock prices in technology, internet, and communications sectors. These studies argue
that the correlation statistic tends to revert towards its long-term average following periods of substantial
deviation.54
A detailed examination of returns indicates that, over time, the US market has not systematically
dominated foreign alternatives.

See, for example, Li, K., & Sarkat, Asani, Should US Investors Hold Foreign Stocks? Current Issues in Economics and
Finance Federal Reserve Bank of New York (March, 2002).
54 See, for example, Brandhorst, Eric, International Diversification, Index Equity Research State Street Global Advisors
http://www.ssga.com/library/resh/ericbrandhorstsinternationaldiversification71502/page.html (11/9/2005). The author
argues that stock return valuation correlation (changes in Price/Earnings ratios) has remained high, but that
fundamental correlation (changes in cash flow growth rates) are relatively low. Even in a new economic regime of
increased globalization, the correlation statistic should, in equilibrium, move closer to its historical average.
53

29

Figu re 2.8: U.S. vs. Foreign Stock Returns

1 97 3 t hrough 2 0 06

M SCI EA FE (Exc luding Japan)

M SCI Japan
S &P 500

2 0.40 %

16.67 %

12 .88 %

1 2.41%

12.02%

1 0.28 %

11.14 %

9.30%

5 .91 %

5.75 %
2.94%
-3 .14 %

1973-1988

1989-2000

2001-2006

1973-2006

Depending on the period under evaluation, foreign stocks have either significantly underpeformed or
outperformed their U.S. large company counterparts. However for the entire period 1973 through 2004, U.S.
and foreign [EAFE] stocks achieved similar compound returns.

INTERNATIONAL BONDS
Prior to the early 1990s, a compelling case could be made for investing in foreign bonds. The U.S. Bond
market ranked in the bottom half of all international fixed income markets five times during the period 1986
through 1993, and never ranked as the top-performing market in any year. Over this period, non-U.S. bonds
outperformed U.S. bonds by 3.37% per year.55 A 1994 study compared U.S. three to five year maturity bonds
with comparable maturity bond indexes from thirteen other nations, from December 1985 to March 1994.56
The study confirmed the poor performance of U.S. bonds compared to the bonds of other developed nations:

Case for International Fixed Income. Unpublished paper by Rowe Price-Fleming International, Inc. Baltimore, MD. For the
longer period March 1973 through December 1988, U.S. bonds (proxied by the Merrill Lynch Domestic Master Index)
earned an annualized compound return of 9%, compared to a return of 11.3% (in U.S. dollars) earned by a composite index
of German, French, Swiss, Dutch, Japanese and British bonds. Rosenberg, M.R., International Fixed Income Investing:
Theory And Practice, The Handbook of Fixed Income Securities, 4th Edition (Irwin Professional Publishing, 1995), pp.10461047.
56 Erb, C. B., Harvey, C. R., & Viskanta, T. E., National Risk In Global Fixed-Income Allocation, The Journal of Fixed-Income
(September, 1994), pp. 17-26. Data for Italy, Spain and Sweden included as of December, 1987.
55

30

Figure 2.9: U.S. vs. Foreign Bond Returns (12-85 - 3/94)

15.7% 15.4%

15.0%
14.3% 14.0%
12.7% 12.1%

11.9%
10.7% 10.1%

9.5% 9.3%
9.0%

en
ed

Sw

ly
It a

in
Sp
a

i tz
er

la

nd

da
an
a

C
S

et

he

er
m

an

s
nd

rl a

gi
u

ce
ra
n

a
m

Be
l

en

Ja

pa

lia
tra
us

rk

6.4%

However, these results can only be evaluated in terms of the risk taken to achieve them. Thus, the higher
returns from foreign bonds over this period must be considered in conjunction with the volatility (risk) of the
foreign markets. Annualized volatility (based on monthly standard deviations of returns measured in U.S.
dollar terms for each national market) during the period under evaluation was as follows:

Figure 2.10: U.S. vs . F oreign Bond Volatility (12/85 - 3/94)

14.9% 15.0% 15.1% 15.4%

15.8% 16.0% 16.1%

13.2% 13.5% 13.5% 13.5%

11.6%
8.5%

nd

K
S

i tz
er

la

n
pa
Ja

m
gi
u

y
Be
l

an

ly

er
m

Ita

s
nd

he
et
N

rl a

ed

en

ce

rk

ra
n
F

en

m
a

in
pa
S

li a
tra
us

an

ad

4.8%

Based on reward to risk it appeared that bonds from any one of these countries would represent a risky
investment for a U.S. investor. However, as we have seen, portfolio diversification benefits are a product not
just of return and risk but of correlation, as well. When asset class correlation is low (or negative), it may be
beneficial to add low return or high-risk assets. Historically, foreign bonds have provided substantial
diversification benefits.
Controversy, however, continues to surround

31

academic opinion regarding the benefits of an

portfolio position in international bonds. For example, Michael Rosenbergs early study comparing bond
markets (1978-1988) indicates that an equally weighted combination of the foreign bond markets resulted in
a portfolio with annual volatility of 5.9%, compared to volatility of 10.6% for a U.S. only bond position.57
Although individual foreign bond markets were volatile in isolation, the pattern of returns differs significantly in
different countries, with upward moves in some national bond markets offsetting downward moves in others.
By 2000, Rosenberg had altered his position on foreign bond investments significantly. In a conference
address to the Association for Investment Management and Research, Rosenberg concluded that a
performance review over the period beginning in 1986 shows that only in the two year window of 1986-1987,
when the U.S. dollar collapsed, did unhedged foreign bonds substantially outperform U.S. bonds.58
Rosenberg concludes that the expected returns from foreign bond investment is about the same as the
expected return from investments in U.S. fixed income investments. Although hedged foreign bonds during the
period 1978 through 1987 exhibited higher returns at a volatility level lower than U.S. bonds, the returns to
U.S. bonds were higher in the subsequent 1988 through 1995 period. Rosenberg points out that, on a go
forward basis, removing the exchange rate risk from foreign bond returns, although positive because it
dampens volatility, may reduce the value of using this asset class because hedging also increases
correlation.59
What, then, is the case for foreign bonds? Although Central Bank policies in most developed nations have
converged in the sense that they formally or informally target inflation, hedged foreign bond positions retain
significant diversification properties. Specifically, they enable investors to gain the advanatage of interest
rate risk diversification without the penalty of exchange rate risk.60 Briefly, the main attractions of Foreign
Bond investments include:
1. Smaller variance compared to investments in the U.S. domestic bond market;
2. Mitigation of interest rate risks that are systematic to domestic investors, but diversifiable to
global investors;
3. Diversifiation of monetary policy mistakes of central banks;
4. Diversification of real economic shocks.61
During the period of the late 1980s through the end of the 1990s the correlation between U.S. and foreign
bonds was relatively constant. Thus, in terms of reducing portfolio risk, the benefits of international bond
diversification appear to offer attractive portfolio construction opportunities in the fixed income area.

Rosenberg, Op. Cit., pp. 111-112. The lack of strong positive co-variance of bond returns during this period was, at least
partially, attributable to heavy market segmentation due to capital flow restrictions and to instability in U.S. monetary policy
during the late 1970s and early 1980s.
58 Rosenberg, Michael R., Foreign Bonds: Not a Strategic Asset, Association for Investment Management and Research
(2000), p. 6.
59 The decision to eliminate exchange-rate risk from a portfolio by establishing a currency hedge remains controversial.
Much academic research suggests that an investor deciding to hedge against currency fluctuations is better served by
hedging foreign bond exposures as opposed to foreign equity exposures. This is especially the case when the primary use
of bonds within the portfolio is for dampening the risk of price declines. See, for example, Winkelmann, Kurt, International
Diversification and Currency Hedging, Modern Investment Management (John Wiley & Sons, 2003), p. 142: Because
currency dramatically affects foreign fixed income, we recommend that investors hedge 100 percent of the currency
exposure in their foreign bond portfolios.
60 Thomas, Lee R., Foreign Bonds: A Strategic Asset, Association for Investment Management and Research (2000), p.
13.
61 Ibid, p. 17: Real shocks affect economies differently. A change in the price of oil does not affect Japan in the same way
that it affects the United States; it will not affect Japanese bonds the way it does U.S. bonds.
57

32

REAL ESTATE
The 1970s and early 1980s were a golden age for U.S. real estate. The high inflation of those years
made real property valuable relative to financial assets. Over a ten-year period, real estate offered high
returns, inflation protection and diversification benefits. Studies indicated that real estate had a negative
correlation to bonds (which lost much of their value in the inflation of the late 1970s), and close to zero
correlation with stocks. Indeed, real estate in this period earned substantially higher returns (3.2% per quarter)
than either U.S. stocks or bonds, yet exposed investors to significantly lower volatility (as measured by
standard deviation of returns of 1.8% per quarter) than either of these asset classes. During the late 1970s,
real estate moved from a trough to a peak, while other capital markets underperformed. Some believed that
real estate was the only asset class suitable for a cautious investor. The following chart, based on a 1986
study covering the fourth quarter of 1973 through the third quarter of 1983, demonstrates the attractiveness
of real estate during this period:62

Figure 2.11: Quarterly Performance Comparison

Q4 1973 thro ugh Q3 1983
9.3%
Quarterly Return
Quarterly Standard Deviation
7.5%

3.2%

1.8%

1.4%

2.1%

1.8%

2.0%
1.0%

0.8%

Real Estate

S&P 500

US Bonds

T-Bills

Inflation

S o ur ce: Downs, D. H., & Hart zell, D. J., Real Est at e Invest ment Trust s, The Handbo ok of Real Estat e Port f ol io M anagement (I rwi n P rof essional
P ubl ishing, 1995), p. 60 1.

The problem with a limited sampling period, however, is that the data may not reflect the true distribution
of long-term returns. Evaluation of data from a later period produces dramatically different results. A
comparative study of returns from an Equity Real Estate Investment Trust index and from the S&P 500
underscores these differences:63

Cited in Downs, D. H., & Hartzell, D. J., Real Estate Investment Trusts, The Handbook of Real Estate Portfolio
Management (Irwin Professional Publishing, 1995), p. 601.
63 Real Estate Finance and Investments, ed. William Brueggerman & Jeffrey Fisher (Irwin Professional Publishing, 1993),
pp. 813-814.
62

33

Fig ure 2.12: Quarterly Perform ance of S&P 500 vs. Real Estate
1986 through 1990
9.33%

Quarterly Return
Quarterly S tandard Deviation

6.24%

3.59%

1.09%

Real Estate

S&P 500

S o u rce : R eal Es t at e Fi nance and I nv es t ment s , ed . W illi am B ruegg erman & J ef f rey F is her (Irw i n Prof es sio nal Publi shing, 1993 ), p p. 813-814 .

An evaluation of real estate using only data from this period would indicate that, in isolation, real estate is
not an attractive investment. However, despite dismal overall performance, real estate still provided some risk
reduction benefits. Specifically, although a large amount of real estate in a stock portfolio would have reduced
returns more than it would have reduced risk (that is, the risk/return tradeoff was not beneficial), the imperfect
correlation of the two asset classes (the equity REIT correlation to the S&P 500 was .807) meant that adding a
small (approximately 20%) real estate position to the portfolio would have created a better overall risk/return
tradeoff.
Until recently, many individual investors portfolios did not include real estate. Private real estate equity
investments in apartments, offices, warehouses, hotels, etc. are expensive propositions that, because of the
necessity for complex financing arrangements, often demand a high degree of leverage. For many individual
investors, either the leverage increases risk beyond their tolerance level, or the collateral and financing
arrangements make private real estate equity investments impractical. Dificulties with real estate investments
include illiquidity (you cannot sell a fraction of a building), lack of marketability (high transaction costs), lack of
geographic diversification, lack of diversification by property type, and high sensitivity to local market financial
conditions including unemployment, and tax policy. Additionally, many investors remember the financial
debacle that followed the S&L crisis, and the collapse of commercial real estate prices as the Resolution Trust
Company dumped property on the market for pennies on the dollar. By the end of the 1980s, many believed
that commercial real estate investment was best left to financial institutions such as insurance companies,
and large pensions and endowments that could afford to own and manage commercial real estate in several
cities or who could purchase private placement debt secured by real estate assets.
Real estate, like many investments, drops in and out of fashion. In the first part of the 21st century, until
the advent of the mortgage liquidity crisis, real estate investors enjoyed a spectacular run up in the price of
single-family homes, as well as solid gains in the
stocks of real estate operating companies, home-

34

building / home-improvement companies, financing companies, and building-supply firms. The investment
vehicle of choice for many individual investors is the Real Estate Investment Trust (REIT). A publicly traded
REIT (that is, a REIT that lists its shares on a stock exchange and, therefore, is both priced and traded like
other stocks) is akin to a closed-end mutual fund. A REITs assets consist primarily of real estate equity
(ownership of properties) and/or debt interests in real estate. Under current tax law, REITs are not taxable
provided that at least 75% of the net assets are invested in real estate related assets, and provided that they
pay out at least 95% of net income to shareholders. REITs are actively managed to increase shareholder value
just as a public corporation is actively managed to promote economic objectives including return on equity,
return on assets, increased market share, and so forth. REITs buy and sell properties like corporations acquire
and spin off divisions; and, like other publicly traded companies, they utilize debt financing. In the period
2000 through 2003, debt for most REITs was in the 30 to 40 percent range.
An Index for REITsthe National Association of Real Estate Investment Trusts, or NAREIT indexstarted in
1978. This is a capitalization-weighted index of all publicly listed REITs; and, as such, is similar in construction
to other capitalization-weighted stock indexes such as the S&P 500 index of U.S. stocks. The NAREIT index
provides the longest pricing history for the asset class of securitized (that is, publicly traded) real estate equity
investments. Today, there are several indexes that track the performance of publicly traded REITs.
In the early 1990s the mutual fund industry launched several real estate equity funds. These funds invest,
for the most part, in stocks of real estate related companies or in REIT shares. Among the early entrants into
the real estate index fund business were DFA (Dimensional Fund Advisors) and Vanguard Funds, each of which
developed a mutual fund designed to track a specific REIT index (e.g. the Vanguard REIT index tracks the
Morgan Stanley REIT index). More recent indexed investments include the Barclays iShare Cohen & Steers
Realty Majors Fund, and the iShares Dow Jones U.S. Real Estate Index Fund.
Investors now have more comprehensive information available to them so that they may better judge the
advantages of adding real estate to an investment portfolio. However, despite the informational advantages,
the benefits of real estate remain uncertain. For example, is real estate an effective hedge against inflation? It
seems that the answer to this question depends on the definition of real estate, on the definition of inflation,
and on the time period under evaluation. The distinction between public and private real estate equity is
important with respect to real estates inflation-hedging ability. Since REIT stocks are stocks of real estate
corporations that own portfolios of assets, a reasonable working hypothesis would be that REITs and private
real estate equity would exhibit similar responses to inflation. However, this is often not the case. One
researcher, for example, notes a positive correlation (+0.41) between inflation and private real estate returns
over the period 1978 through 1997; but argues that the value of the correlation statistic between inflation and
securitized real estate is 0.00 for the same period. A correlation statistic is only an average for the entire
period; and, examination of sub-periods can reveal return patterns significantly different from the overall
average. For example, although the study concludes that private real estate provided a meaningful positive
inflation hedge, this is not to be expected under all market conditions: when space markets experience
significant excess supply, as in the 1988-1992 period, the presence of unanticipated inflation will not
necessarily result in a rise in real estate returns.64
Other than high expected returns, academic studies offer several reasons for inclusion of real estate
securities in a prudent and balanced investment portfolio:

Real estate may offer a hedge against inflation; and,

Sanders, Grayson, An Updated Look at Asset Allocation: Private and Public Real Estate in A Multi-Asset Class Portfolio,
The Real Estate Finance Journal (Winter, 1998), pp. 5-13.

64

35

Real estate securities have low correlation with other common stocks.65

The chart below shows how real estate has behaved vis--vis inflation and the S&P 500 from 1978
through 2006.

Fi g 2.13: Re al Estate , In flatio n and S& P 500

3.0%

R ol ling 36 Month R eturns ex press ed as monthly fig ure, 1 978 - 2006

2.5%

2.0%

1.5%

1.0%

0.5%

0.0%

-0.5%
S& P 500
DJ Wilshire Real Estate Securiti es
-1.0%
U.S. Inflation

-1.5%
1981

1983

1985

1987

1989

1991

1993

1995

1997

1999

2001

2003

2005

By combining asset classes that respond differently to future economic conditions, the portfolio becomes
more stable and, therefore, less likely to produce unacceptable downside returns. Susan Hudson-Wilson, for
example, argues that real estate is a good portfolio diversifier.66 After developing a custom index reflective of
the four quadrants of real estate investing (public and private debt instruments / public and private equity
positions), she notes that the value of the correlation statistic between stocks (S&P 500) and real estate is
+0.547 during the period 1987 through 2000, and the correlation between bonds (Lehman Corporate /
Government Bond Index) and real estate is +0.284 during the period. She notes: when the return to an asset
class is high enough, or the risk is low enough, and/or the correlation reflects a sufficiently different pattern of
returns, the asset class earns a place in the portfolio for at least a portion of the return-risk spectrum. Real
estate meets these tests, and is therefore a component of the well-diversified mixed-asset portfolio.67

A comprehensive discussion of real estates role in the investment portfolio appears on our website at
http://www.schultzcollins.com/files/IQ2004Q2.pdf.
66 Hudson-Wilson, Susan, Why Real Estate?, The Journal of Portfolio Management (Fall, 2001), pp. 20-31.
67 It is interesting to note that, when the four quadrants are combined into a capitalization-weighted index of real estate
debt and equity securities, the index suffered no nominal dollar losses in any year from 1982 through 2000. However, this
observation must be tempered by noting that, for most of this period, the index is weighted primarily to private debt and
equity.
65

36

Questions remain, however, regarding the optimal portfolio allocation to real estate. Susan Hudson-Wilson
suggests that real estate is a risk-reducer at low to moderate risk and return levels, and so has no role in
highly risk-tolerant portfolios. She pegs the optimal allocation for very risk-averse investors at 27%; however,
this weight drops to zero rather quickly as one moves up the risk spectrum seeking higher returns. Thus, in her
opinion, real estate is primarily suitable only for investors interested in capital preservation.
Ziobrowski, Caines & Ziobrowski arrive at exactly the opposite conclusion: conservative managers seeking
low risk and willing to tolerate lower returns should hold little or no real estate. Managers seeking higher
returns who are more tolerant of risk should hold some real estate, but not very much, ranging between 4%
and 18% of the total portfolio maximum.68
Grayson Sanders advances the proposition that the optimal portfolio turns out to be 40 percent bonds,
30% stocks, and 30% public real estate. This is probably not a feasible solution in the marketplace because of
the mismatch with the size of the investable universe. From a practical perspective we can take comfort from
this analysis that a 10 to 15 percent allocation to either public or private real estate or a combination thereof
can be readily justified.69
The diversity of learned opinion is a good example of how statistical conclusions are hypersensitive to the
sampling period and to the way the variables of interest are defined. The investor must clearly understand the
role of real estate within the portfolio. It is not a guaranteed safety net against the ravages of unexpected
inflation, nor is it an asset class that will produce returns with smoothed volatility. Allocation weightings
reflect precise calculations, investor preferences, and common sense. If an investor holds no investment real
estate assets outside of the portfolio, a modest allocation to real estate appears to be prudent and suitable.
However, if an investor owns a large amount of private real estate equity, a heavy allocation towards real
estate within the investment portfolio may create redundancies and unnecessary asset concentration risks.

EMERGING MARKETS
Emerging markets are the equities markets of countries still in the early decades of forming the
institutions that make for economic success. In emerging markets, fundamental elements of a market system,
such as property rights, contract law, freedom of speech, and due process, may not be securely instituted. The
stock markets of developing nations can be extremely volatile. For example, in 1990, the Taiwanese Stock
Exchange Index started the year near the 5,000 level. By the end of the first quarter, it had reached 12,600;
six months later it had plunged to 2,500.
A broader perspective on emerging market volatility is provided by a BARRA study of the period 1987199170. In general, stock market volatility (as measured by standard deviation) was considerably higher for
emerging market economies than for the U.S.:

68 Ziobrowski, A.J., Caines, Royce & Ziobrowski, B. J., Mixed-Asset Portfolio Composition with Long-Term Holding Periods
and Uncertainty, Journal of Real Estate Portfolio Management (Vol. 5, 1999), pp. 139-144.
69 Sanders, Grayson, An Updated Look at Asset Allocation: Private and Public Real Estate in a Multi-Asset Class Portfolio,
Real Estate Finance Journal (Winter, 1998), pp. 5-13.
70 Divecha, A .B., Drach, J., & Stefek, D., Emerging Markets: A Quantitative Perspective, The Journal of Portfolio
Management (Fall, 1992), p. 42.

37

Figure 2.14: Equity Volatility in Emerging Markets & the United

S tates
1987 through 1991
Country
Argentina
Turkey
Brazil
Taiwan
Portugal
Greece
Mexico
Indonesia
Korea
US

Standard Deviation
108%
84%
74%
63%
61%
56%
56%
39%
29%
17%

Although returns from emerging market investments were also high during this period (19.7% vs. 12.6%
for the London Financial Times World Index of developed nations), the variability of returns from individual
nations is striking. A more recent survey of emerging market returns, covering the period 1975 through 1999,
derives similar results for the longer-term series of monthly returns:

Figure 2.15: Summary Statistics for

Emerging Markets 1975 - 1999
Country

Return

Standard
Deviation

Argentina

4.54

25.86

Brazil

2.14

16.99

Chile

2.46

10.52

Colombia

2.07

9.20

Greece

1.20

9.98

India

1.41

8.05

Indonesia

2.08

17.56

Jordan

0.78

4.77

S. Korea

1.61

11.17

Malaysia

0.91

10.51

Mexico

2.04

12.24

Nigeria

1.81

12.56

Pakistan

1.00

8.98

Philippines

2.35

11.23

Portugal

2.41

11.09

Taiwan

2.27

13.34

Thailand

1.35

10.08

Turkey

3.83

20.04

Venezuela

1.89

14.72

Zimbabwe

1.08

10.43

The average monthly return is 1.62% over the

38

period under evaluation. This return compares to

returns of 1.26% for EAFE and 1.36% for the U.S. When each country is considered in isolation, monthly
volatility for the period averaged 12.46%--a number evidencing significant investment risk. However, when
considered as a single portfolio, an investment in the aggregate emerging markets asset class had a monthly
volatility of only 5.70%. This compares to monthly volatility of 4.73% for EAFE andf 4.23% for the U.S.
The more recent data appears in a study by Conover, Jensen, and Johnson. 71 The authors examine the
behavior of emerging markets over the period 1976 through 1999, including all emerging markets with at
least ten years of returns data as of December 1999. They calculate returns for a composite emerging markets
index, with underlying countries weighted according to each markets gross domestic product. Thus, the larger
economies of Brazil, India and Mexico are heavily weighted in the index, while Zimbabwe contributes little.
They note: The US market offers a substantially higher return for a given level of risk, which indicates that
emerging markets are not attractive stand-alone investments. But, in the authors opinion, this does not rule
out emerging markets as elements of diversified portfolios; the attraction of the emerging markets lies to a
large extent in their much lower average correlations with developed markets.
The following graph illustrates the ending value of $1,000 invested in various indices over the period 1988
through 2006. It is noteworthy that the emerging markets asset class has performed extraordinarily well
during 2004 and 2005 as the US Federal Reserve engaged in a series of interest rate increases.
Figure 2.16: Ending Value of $1,000 in U.S. and F oreig n Stock s
$1 6,0 00

198 8 through 2006

MSCI Emerging Composite

$14,765.17

S&P 500

$1 4,0 00

MSCI EAFE
$1 2,0 00

MSCI Emerging Composite

$1 0,0 00

S&P 500
$8,844.38

Blended Index of 50% S&P

500/25% EAFE/25% Emerging

Blended Index
$7,683.47

$ 8,0 00

$ 6,0 00
MSCI EAFE
$4,016.9 0

$ 4,0 00

$ 2,0 00
$ 1,0 00

1988

1990

1992

1994

1996

1998

2000

2002

2004

2006

One source of diversification benefit lies in the fact that, unlike stocks of other developed economies,
performance of emerging market stocks is not closely correlated with US Federal Reserve policy. This implies
that developing countries are less likely to establish monetary policies that align with those of the developed
countries. Thus, apart from the compensation of the raw investment returns they offer, the risk of investing in
emerging markets may also be compensated by a reduction in overall portfolio volatility during periods of
restrictive U.S. monetary policy (that is, tight money brought about through increases in the Federal Funds
discount rate). While the stock returns of developed nations seem to correlate closely with the performance of
U.S. stocks over interest rate cycles, the returns from emerging markets evidence a greater degree of
71 Conover, C. Mitchell, Jensen, Gerald R., & & Johnson, Robert J.,. Emerging Markets: When are they Worth It? Financial
Analysts Journal, (March/April, 2002), pp. 86-95.

39

independence. The authors conclude: Even for investors interested in maintaining relatively low-risk equity
positions, the optimal portfolio has a large international exposure and most of it is in emerging market stocks.
The Conover, Jensen & Johnson study echoes conclusions made by previous studies:

Returns of emerging market nations are not strongly correlated, so investment in the overall asset
class is substantially less risky than an investment in any single individual emerging market; and,

Returns of emerging market nations are not strongly correlated with those of developed nations;
thus, they offer an opportunity for effective portfolio risk reduction.

The low correlation of returns means that, in a well-diversified portfolio of emerging market investments,
the pattern of gains or losses in one nations markets offset those in others. Several studies demonstrate that
adding emerging markets to an international portfolio reduces overall portfolio risk. Divecha, Drach, & Stefek
reviewed varying mixes of the Financial Times World Index with the International Finance Corporation Emerging
Market Index. They concluded that until reaching about 20% in the Emerging Markets Index, the risk of the
overall portfolio decreases, because of the low correlations between the two.72 Using another set of
benchmarks (Morgan Stanley Capital World Index and Genesis Emerging Markets Fund) in a study of the
period July 1989 to June 1993, J. Paulson-Ellis demonstrated that an investor can reduce risk until 40% of
the portfolio is in emerging markets.73 Finally, Conover, et. al, concluded that even low-risk portfolios can
accommodate up to 33% in emerging market equities.74 Thus, it seems clear that emerging markets offer
important diversification benefits to investors wishing to implement prudent investment portfolios.75
Although the above discussion covers asset classes commonly used to construct investment portfolios, it
is not an exhaustive list. The SCLC website offers a more technical discussion of the benefits of other asset
classes, including inflation-adjusted bonds and commodity investments. The working paper entitled Investing
in Commodities: Issues and Current Research is particularly useful.76

Op. Cit., Divecha, Drach & Stefek, p. 49.

Paulson-Ellis, J. Introducing Emerging Markets, Managing Emerging Market Portfolios (AIMR, 1994), p.15.
74 Op. Cit., Conover, et al., p. 92.
75 An in depth discussion of the role of emerging markets in an investment portfolio is found on the SCLC website at
http://www.schultzcollins.com/files/IQ2004Q1.pdf.
76 This is found at http://www.schultzcollins.com/files/Investing%20in%20Commodities.pdf.
72
73

40

CHAPTER 3: DIMENSIONS OF RISK AND RETURN

ASPECTS OF THE U.S. MONEY MANAGEMENT INDUSTRY

Investors need to develop expectations regarding the risks and returns of investment choices. In many
respects, the problem of portfolio construction is the problem of
1. Selecting assets under conditions of uncertainty; and,
2. Appropriately weighting of each investment within the portfolio.
As suggested earlier, large segments of the retail investment industry (stockbrokers, financial planners,
money managers) focus on the security selection decision. The quest is to discover securities that are
mispriced (i.e., undervalued) and to form portfolios that are collections of stocks and bonds that, in the
managers opinion, offer the best opportunities for positive investment returns over the forthcoming period.
Although money managers use many terms to catagorize their investment philosophies (momentum,
contrarian, market-neutral, sector rotational, etc.), two terms are in common usage. These are valueoriented managers and growth-oriented managers. Value and Growth investment styles reflect the fact that
certain categories of stocks share specific characteristics and exhibit common return patterns. These stock
categories, moreover, perform differently than stocks grouped in other categories. The common or shared
characteristics are, in the parlance of economists, factors. Stocks sharing high exposure to certain factors
are more likely to exhibit relatively high correlation. The returns to a portfolio of large company stocks, for
example, tend to differ from the returns to a portfolio of small company stocks because each group is exposed
to differing risk factors. Companies with small capitalization, for example, have greater exposure to
bankruptcy risk during recessionary periods. Thus, economist distinguish between return-to-small and
return-to-large risk factors. Likewise, there exists return-to-value and return-to-growth risk factors.
In general, Growth-oriented investment managers form portfolios of companies that:
1. Demonstrate above-average values in financial ratios focusing on sales growth, revenues and
profits, return on assets, and return on equity; and,
2. Offer solid competitive advantages in the marketplace (i.e., patent protections, well-known
brands, ownership of intellectual property, and so forth).
Although the share prices for these companies may be high, financial analysis may predict continued
above-average growth in both marketshare and return on the firms invested capital.
However, the risks associated with growth investing are both obvious and notorious:
1. Investors may fall into the trap of overpaying for attractive or glamour companies with the result
that even though the company continues to produce stellar accounting statements, the return to
the investors is less than expected; or,
2. Marketplace shocks may reduce rock-solid competitive advantages as new technologies leapfrog
over existing business systems, as new competitors suddenly announce their entry into the
marketplace, as new tax law or regulatory restrictions change the profitability of a business, as

41

new mergers and acquisitions detract from core business activities, and so forth.
Growth stock companies tend to be highly profitable and often exhibit rapidly increasing earnings and
market share. The high prices paid for many technology, computer and communications stocks in the late
1990s represent a recent highwater mark in the popularity of growth-style investing.
Value-oriented investors generally purchase securities when the market value of their shares approaches
the book value listed on their financial statements. Relative to growth stocks, value stocks tend to be less
profitable and to exhibit less robust opportunities for future earnings growth. The universe of value investors,
however, includes several well-known names including, Benjamin Graham, Warren Buffet, and John Neff.
Graham, for example, developed a formulaic, yet rigorous, approach to stock picking. He advised investors to
avoid stocks when their normalized [(current price average price) standard deviation of price] price to
earnings ratio is above 25. Many of Grahams market timing and stock selection formulae first published in
the 1930s have not stood the test of time. However, several formulas that reflect modern styles of portfolio
formation through stock picking still resonate strongly: when the price of a stock is less than 1.3 times the
tangible book value, it should be a good value for the investor.77 Graham cautions investors that it is better to
pay too little for a company with ugly financials than to pay too much for a company with attractive financials.
Although Graham focused more on a method for picking individual stocks as opposed to a method to
understand and manage assets within a portfolio context, nevetheless, his research offers useful insights; and,
it continues to command respect from financial analysts.
In Grahams world, investors fall into two groups: (1) projection investors; and, (2) protection investors.
One commentator updates Grahams distinction as follows:
Projection investors want to find the next Microsoft Corporation. If they can find the next
Microsoft, they do not care how many 3Coms or Quarterdeck Softwares or Ashford.coms they
end up with. As long as they get one Microsoft, they will achieve their goals. In contrast,
protection investors want to make certain that they do not get wiped out. They are not
concerned about being right only once. They want to minimize the number of times and the
consequences of being wrong.78
Projection investors wish to hit a home run. At the extreme, they are treasure hunters searching for the
next glamour stock. The protection investor, by contrast, avoids overpaying for hype, trends, fads, and other
ephemeral characteristics that divorce a securitys market price from its fundamental value. In Grahams
view, the stock market is not efficienti.e., may not price securities close to their fundamental value.
Graham encourages investors to develop sufficient information to allow them to identify a subset of mispriced
securities and to invest therein.
Such a recommendation, however, cannot comfortably co-exist with the hypothesis of market efficiency
(markets are efficient if the costs of acquiring information offer, on average, no ability to make abnormal
profits based on such information). Although the efficient market hypothesis concedes that investors will over
react with enthusiasm for some stocks or under react for other stocks, it is impossible to determine ahead of
time which way the crowd will react to news about any specific security; and, therefore, it is impossible to
develop any predictive system for profitable stock selection.
The projection/protection [growth/value] investor distinction leads to another important observation. The
projection investor, seeking the next ticket to fame and fortune, avoids diversification. This is often the case
despite the fact that the treasure-hunting investing style cries out for a safety net: diversification is probably
more important for projection investors than they might realize: The odds of missing the next Microsoft are
Zweig, Jason, Lessons and Ideas from Benjamin Graham, Equity Analysis Issues, Lessons, and Techniques (AIMR,
2004), pp. 9-17.
78 Ibid.
77

42

high, and a growth portfolio with only a few stocks has a high likelihood of being wiped out.79 Interestingly,
however, value stock pickers who believe that they have not overpaid for a security (i.e. that they have found
good value in the marketplace), may also eschew diversification because they believe that they are less
vulnerable to the spectacular crashes of high-flying, high-priced stocks. Warren Buffet, for example, has a wellknown disdain for diversified portfolios, and preaches the gospel of finding good companies at a good price
and holding them for a length of time sufficient to reap the expected rewards. In the Buffet ideology, it does
not matter which way the crowd will go as long as the investor has the conviction that his valuation analysis
remains correct; and has the patience to hold the stock through the ups and downs of market irrationality.
John Neff, the famous portfolio manager of the highly successful Windsor fund (a fund that picked stocks
according to a philosophy that is a comfortable fit with Grahams), recently replied to the question How many
stocks should a portfolio hold? as follows: With Windsor, we probably had 60, and the top 10 were about 40
percent of our portfolio.80 The titans of the money management industry from both the growth and value
stock picking camps embrace asset concentration risk as a reasonable investment strategy.81

VALUE/GROWTH ASSET CLASS INVESTING V. UNDERVALUED STOCK PICKING

In 1982, Tom Peters and Robert Waterman published a best-selling study of major American companies
entitled In Search of Excellence.82 Among the criteria they used to identify excellent companies were six longterm (1961 through 1980) measures of financial superiority:
Compound growth of assets;
Compound growth of shareholder equity;
Average ratio of market value to book value;
Average return on total invested capital (net income divided by total invested capital);
Average return on equity; and,
Average return on sales.
A 1987 study by Michelle Clayman took In Search of Excellence as its departure point. She hypothesized
that an investor might be able to realize superior returns by holding a portfolio of companies that have
demonstrated financial excellence. She tested the hypothesis by screening the S&P 500 for companies with
Ibid.
Neff. John B., A Conversation with Legendary Value Investor John B. Neff, CFA, Equity Analysis Issues, Lessons, and
Techniques (AIMR, 2004), pp.5-8.
81 It is interesting to note that there is common ground between Wall Street gurus and the advocates of the efficient market
hypothesis (EMH). Both camps can agree that if the investor has no forecasting skill whatsoever, it makes sense to
diversify the portfolio completely lest stupid stock picks destroy wealth. Conversely, if one has perfect forecasting skill (i.e.,
a prophet of profits), it makes sense to avoid diversification and buy the single investment that will generate the highest
return over the forthcoming period. Any other course of action would simply be a waste of money. Thus the degree of
asset concentration risk to which it is reasonable and prudent to expose investment wealth is a function of forecasting skill.
But forecasting skill is a measurable talent and is subject to statistical verification. Thus, the degree of prudent
diversification rests on a rigorous quality control testing and evaluation system that measures the prevalence and
magnitude of forecasting errors within (i.e. internal to) the money management organization. Evaluation of internal
forecasting skill is not the same as evaluation of the money managers external track record although success in one area
should manifest itself in success in the other. EMH proponents point out that it is difficult to demonstrate a degree of
forecasting accuracy that justifies picking just a few stocks. For example, if the opportunity set of investments consists of
the S&P 500, what level of demonstrable forecasting accuracy is required to justify exposing wealth to the ups and downs
of, say, only 20 stocks? Currently, this is a topic of great interest across the academic, money management, and litigation
communities.
82 Peters, Thomas J. & Waterman, Robert H., In Search of Excellence: Lessons from Americas Best-Run Corporations,
1982.
79
80

43

the best overall accounting ratios. She further screened the S&P 500 for unexcellent companies that could
act as a control group. Accounting ratio values for the excellent and unexcellent companies for the period prior
to the formation of the portfolios were as follows:

F ig u re 3. 1 : E xc e llen t vs . Un e xc elle n t Co mpa n y Ra tio s:

1 97 6 - 19 8 0
25
20
15
10
5
0
A s s et
Gr o w th

Eq u ity
Gr o w th

Ma r ke t- Bo o k
Ra ti o

E xc e lle n t C om pa nie s

Re tu r n o n
To ta l C a p ita l

Re tu r n o n
Eq u ity

R e tu rn o n
Sa l e s

U ne xce lle nt C o m p an ie s

Clayman tracked results from equally weighted portfolios of excellent and unexcellent companies from
1981 through 1985. As expected, the excellent company portfolio outperformed the S&P 500 index by 1.1%
per year over the period. The portfolio of unexcellent companies, however, outperformed the S&P 500 by
12.4% per year.83 Clayman had discovered, for this five-year period, that stocks of good companies were not
necessarily good investments.
Initially, many observers argued that Claymans findings represented a mere statistical anomaly for the
brief period under evaluation. However, Nobel Prize winner William Sharpe, in a well known study of mutual
fund performance,84 concluded that the returns of U.S. equity mutual funds differ from each other along two
key dimensions: One [dimension] may loosely be termed value/growth, the other small/large. Eventually,
Sharpes terminology (growth and value) replaced Claymans terminology (excellent and unexcellent).
Sharpe published his study only a few months after a separate and independent historical review of U.S.
stock returns by two professors of economics: Eugene Fama and Kenneth French.85 Fama and French
concluded that the primary determinants of U.S. stock returns were subsumed in three identifiable factors:
1. The sensitivity of the stock to systematic market risk (beta, in the lexicon of financial
economics);
2. A value/growth factor; and,
3. A large company/small company factor.

Clayman, Michelle, In Search of Excellence: The Investors Viewpoint, Financial Analysts Journal (May-June, 1987), pp.
54-63.
84 Sharpe, William F., Asset allocation: Management Style and Performance Measurement, The Journal of Portfolio
Management (Winter, 1992), pp. 7-19.
85 Fama, Eugene F., & French, Kenneth R., The Cross-Section of Expected Stock Returns, Journal of Finance (June,
1992). Pp. 427-465.
83

44

The Fama/French study documents the relative superiority of returns of unexcellent companies during the
period July 1963 through December 1990 for both U.S. large company stocks and U.S. small company stocks.
Additionally, in 1993, Sharpe published research into the stock markets of major Asian and European nations
(over the period January 1981 through June 1992). The study demonstrated that the relative superiority of
unexcellent firms was a worldwide phenomenon.86 As noted, Sharpe calls stock of unexcellent companies
Value Stocks, while those of excellent companies he calls Growth Stocks. These are now the standard terms
for the two investment styles. In a nutshell, growth style investors seek to buy the securities of companies with
substantial growth prospects in the hope that superior corporate growth will provide high returns to investors
over the long run. Value style investors seek to buy securities that can be purchased for prices that are low
relative to the companies estimated underlying values.87
Rather than using fundamental analysis (close examination of corporate financial performance data) to
find a few stocks offering good value, money managers, by the early 1990s, formed broad-scope portfolios by
dividing the marketplace into value and growth stock asset classes. Traditionally, investors built portfolios
using only three building blocks (stocks, bonds, and cash). The expansion of asset classes along value /
growth, and large / small market capitalization dimensions, however, suddenly increased the number of
building blocks for investors. Portfolios now could incorporate allocations to large and small stocks, growth
and value stocks, foreign and domestic stocks, and so forth. Each asset class represented return-generating
factors (style factors, capitalization factors, etc.) in the marketplace. Depending on their specific investment
objectives, risk tolerance and planning horizons, by the mid 1990s investors sought to customize portfolios by
loading either for or against these factors. Single factor portfolios (owning some combination of stocks and
bonds) became multifactor portfolios.88

Capaul. C., Rowley, I, & Sharpe, W. F., International Value and Growth Stock Returns, Financial Analysts Journal
(January/February, 1993), pp. 27-36.
87 Ibid., p.27. A recent monograph by New York University professor Aswath Damodaran, a noted expert in the subject of
business valuation, considers the question of how strong corporate management and superior past earnings growth affects
the prospects for future stock price increases: the earnings announcements of firms are judged against expectations
and the earnings surprise is what drives prices. You would therefore not expect any correlation between the magnitude of
the economic value added and stock returns or even between the change in economic value added and stock returns.
Stocks that report the biggest increases in economic value added should not necessarily earn high returns for their
stockholders. Interestingly, Damodaran extends Claymans research in a study that ranks all stocks rated by Standard
and Poors according to quality ratings based on quantitative financial measures and qualitative evaluations of corporate
management. During the period under evaluation the lowest rated stocks had the highest returns and the highest rated
stocks generated the lowest returns. However, forming portfolios based on the Fortune magazines list of fifty most
admired companies (1983 through 1995) resulted in substantial outperformance compared to portfolios of the least
admired companies. Damodaran, Aswath, In search of Excellence! Are Good Companies Good Investments, (New York
University website, 2005), p. 136.
86

The most famous of the single factor market models is the Capital Asset Pricing Model (CAPM) developed independently
by William Sharpe and others in the 1960s [Sharpe, William F., Capital Asset Prices: A Theory of Market Equilibrium under
Conditions of Risk, Journal of Finance (September, 1964), pp. 425-442]. The CAPM measures the relationship between
the rewards that an investor might reasonably expect in terms of the risk that the investor assumes. In equilibrium, the
following equation describes the expected return on any security:
88

Expected Return of Security A = (Price of Time) + (Price of Risk)(Amount of Risk)

where:
The Price of Time is the minimum return required to compensate the investor for investing rather than consuming [this is
the T-Bill or risk-free rate which also incorporates the expected rate of inflation];
The Price of Risk is the return offered in the marketplace that is in excess of the risk free rate [this is the risk premium];
and,
The Amount of Risk is the volatility of the stock relative to the volatility of the market as a whole [this is the stocks Beta].

45

THE HISTORICAL EVIDENCE: IS VALUE A STRONG LAW OF ASSET PRICING?

Things became a bit confusing as academic research continued to explore the value phenomenon. What,
for example, is a value stock? Some argue that high dividend paying stocks are value stocks because the
payment of cash to investors constrains future corporate growth by inhibiting the firms ability to fund future
projects. Others argue that highly regulated companies such as utilities are value companies because of
regulatory constraints on their ability to set rates in the marketplace. Stilll others argue that many nondividend paying companies are value stocks because their low share prices indicate the need to conserve cash
to meet current obligations (i.e., the firms have high bankruptcy risk). One of the complexities faced by
investors is the plethora of investment products that carry the label value but that utilize significantly different
definitional criteria for security selection.
By the mid 1990s, however, it appeared that the return advantage of value stocks was substantial. For
example, a comparison of the Growth, Value, and S&P 500 Stock indexes (maintained by the Center for
Research in Securities Prices) reveals the following results for an initial investments of $1,000:

The CAPM uses the stock market as a single factor to determine the amount of risk, and, hence, the amount of expected
return for any investment. Thus, Sharpes model is also known as a single-factor asset-pricing model. A stock earns more
than the risk-free rate only because it takes on some additional amount of stock market risk. Another way of saying this is
to state that market risk [Beta], under the CAPM, is the only priced risk factorthe single factor that will generate an
expected reward over and above the risk-free rate. If you dont take market risk, all you can expect to earn is the risk free
rate.
Although CAPMs central hypotheses [return is a function of a single risk factor which, in turn, can be proxied by the stock
market] may seem counterintuitive, Sharpe developed an analytically derived CAPM based on rigorous mathematics. He
asked what the consequences for asset pricing might be if the expected rewards for individual stocks were not calibrated
or risk-adjusted to the rewards offered by the market in general. The mathematics lead to a theory of arbitrage that
assures that asset prices adjust according to the predictions of CAPM. If General Motors offers a better reward-to-risk ratio
than the market in general, then investors will sell their positions in the general stock market (lowering market prices) to
buy shares of GM (raising the price of GM). This process will continue until such time that the two are in equilibrium with
respect to both their prices and their risks. This process happens quickly (people do not want either to leave money on the
table or to remain in inferior investments) with the result that, for the most part, the risk-adjusted price of any investment is
rapidly calibrated to the risks and rewards of the general market. There is little possibility of finding a mispriced security;
or, in Benjamin Grahams words, a stock that offers a superior investment value.
Following Sharpe, economists wished to know if there were factors other than the market that could explain the return
generating process for securities. If other factors could be identified and measured, portfolios might be formed to capture
the risk premiums from a variety of factor exposures. Factor analysis might lead to better measurement and management
of investment risk as well as to higher expected investment returns. The hunt for priced risk factors began; and the
multifactor investment portfolio was about to be born.

46

Figure 3.2: Ending Value of $1,000

Large Company Growth vs. Large Company Value Stocks
1973 thro ugh 2006
$1 0 0,0 00

Large Company Value Index*

$77,802.31

$ 5 0,0 00

S&P 500
$ 36,222.68

$ 3 0,0 00
$ 2 0,0 00

Lar ge Company Grow th

Index**
$22,647.97

$ 1 0,0 00
$ 5,0 00
$ 3,0 00
$ 2,0 00
$ 1,0 00
$5 00
$3 00
$2 00
$1 00

1973

1976

1979

1982

1985

1988

1991

1994

1997

2000

2003

2006

* F ama-French Large Value Index

** Fam a-French Large Gro wth Index

Not only did the studies published in the early and mid 1990s note the higher return to value, they also
struggled to explain why diversified portfolios of value stocks do not manifest relatively higher risk (that is,
standard deviations). Empirical performance results suggest the possibility that portfolios with value tilts offer
a free lunch to investorsmore return with less risk. The attractiveness of a tilt toward value style investing
as a portfolio risk-control mechanism seems pronounced:

47

F i gu re 3 . 3 : Gr ow t h v s. V a lu e
1 973 th ro u gh 2 006
A n nu aliz e d R et u rn
S t an da rd D e viat io n
19 .2 1%

1 8.7 5 %
1 3.6 6 %
9 .6 1%

Larg e Value Ind ex *

La rge Gr ow th Inde x* *

* F am a-F re nc h La rg e V a lue In dex

* * F am a-F ren c h Large G ro w t h I ndex

During this period, value style investing in large U.S. company stock earned a higher return at a lower level
of risk than growth style investing. On average, stocks of firms with good earnings growth prospects yielded
poorer returns than stocks of firms with distressed earnings. In explaining this phenomenon, one popular
theory was the cost-of-capital argument:
Companies experiencing earnings distress have higher average stock returns than companies
with good earnings prospects, suggesting that the market demands compensation for this
risk factor.companies with poor earnings prospects pay higher costs of capital than
companies with good earnings prospects. When they borrow money at a bank, they pay higher
interest rates. When they issue stock, they receive less money. Since they receive less money
from a stock issue, their expected returns are higher. The fastest-growing stock returns are
likely to come from investing in slow-growing companies.89
The academic community was split along several lines:
1. Value investing is a strong law of asset pricing. The return to value stocks was
explainable either by a cost-of-capital argument or by a priced risk factor argument.
The priced risk factor argument claimed that value stocks were risky but that the risk
(as measured by historical standard deviation) had not yet manifested itself in the
marketplace. Skeptics began to refer to the metaphysical risk dimensions of value
investing.
2. The returns to value style investing represents agency risk.90 Decision makers for
pension plans, endowments and other institutional investors must report to boards
of directors and must justify their investment decisions. The prices of value stocks
are relatively depressed because institutional decision makers are more comfortable
holding a blue-chip stock portfolio of good companies despite the fact that value
89 Booth, David B., Growth Stocks: Earnings Growth Vs. Stock Returns, (May, 1994) Dimensional Fund Advisors, Inc., pp.
1-2.
90 Lakonishok, J., Shleifer, A., & Vishny, R. Contrarian Investment, Extrapolation, and Risk, Journal of Finance (December,
1994).

48

stocks may have higher expected returns. This reluctance to hold value stocks
means that the price of such stocks adjusts downward to reflect the demand-to-hold
schedulethat is, value stocks become a relative bargain.
3. The return to value stocks is not a strong law of asset pricing. Rather than
representing a priced risk factor, the value phenomenon represents either a striking
form of data mining (in which case the return to value should not be expected to
continue)91 or proof that the markets are not efficient (in which case the return to
value is a pricing anomaly that could be exploited by investors for their benefit).
Empirically, if we proxy the growth segment of the U.S. stock market by the Fama-French Large Growth
Index and the value segment by the Fama-French Large Value Index, the comparative performance of value vs.
growth during the period 1989 through 2006 is as follows:

Figure 3.4: Value Premium Over Growth

(annual differential 1973 - 2006)
27.08%
21.46%

19.93%

18.12%
11.10%
5.84%

6.08%

16.89%
12.25%

6.19%

7.44%
0.52%

-3.52%

12.38% 13.73%
8.78% 9.35%

0.27%

-0.89%
-6.41%
-10.19%

-18.75%

19.43%
14.41%

17.16%
17.13%

13.43%

-13.81%
-15.98%

-7.73% -7.90%
-11.03%
-18.41%

-29.65%

More recent academic research treats the value/growth debate as part of a general examination of
portfolio factor analysis. This analyslis asks:
1. Which factors are true determinants of return;
Knez, P.J. & Ready, M.J., On the Robustness of Size and Book-to-Market in Cross-Sectional Regressions, Journal of
Finance (1997), pp. 1355-1382, suggest that the statistical significance of the value premium depends on the one percent
of extreme observations. Loughran, T., Book-to-Market across Firm Size, Exchange, and Seasonality, Journal of Financial
and Quantitative Analysis (September, 1997), pp. 249-268 argues that the value premium is limited to small cap stocks
and is primarily an artifact of tax loss-harvesting strategies (the January effect). See also, Black, Fischer, Beta and
Return, Journal of Portfolio Management, (Fall, 1993), pp. 8-18.
91

More recently, Dhatt, M., Kim, Y. & Mukherji, S., The Value Premium for Small-Capitalization Stocks, Financial Analysts
Journal (September/October. 1999) test stock returns over the period July 1979 through June 1997. They contend that
the value premium was pervasive throughout the entire year (i.e., it is not an artifact of a January effect), and that it is
both statistically and economically significant. Additionally, after controlling for small, thinly traded stocks, their tests
indicated no substantial diminution in the value premium. These studies highlight the continuing academic controversies
surrounding value style investing.

49

2. What is the expected compensation for exposing portfolio wealth to the factors (e.g., what is the
expected premium for assuming value stock risk); and,
3. What is the optimal degree of exposure to the relevant factors (that is, what is a prudent asset
allocation)?
Having identified priced factors and having forecasted their expected reward, investors can determine
the extent to which they wish to tilt their portfolio to capture the expected rewards or to limit overall portfolio
risk.
Modern factor analysis follows three broad areas of research:
1. The factors that determine portfolio returns are fundamental attributes of individual firms
market, industry or accounting characterisics;
2. The factors that determine asset prices are identificable marcro economic forces such as
inflation, growth in the gross national product, interest rates, energy costs, unemployment, etc.;
or,
3. The factors are not able to be specified (i.e., unobservable factors that are implicit to the matrix of
time series returns). However, once the unobservable factors are found, they may be retranslated into known factors through various advanced statistical techniques.
The crux of the matter centers on whether the value premium represents a priced risk factor. If the
answer is in the affirmative, this constitutes a strong rationale for including a value tilt in the portfolio. If the
answer is in the negative, the justification for including a value tilt rests primarily on the probability, based on
empirical observations only, of a successful investment outcome. The lack of consistency in the value
premium suggests that investors cannot consider it to be a strong law of asset pricing. The more recent
comments of Fama and French indicate that they consider that their research is primarily helpful as a tool for
describing a set of investment opportunities richer than heretofore described under the more classical asset
pricing models. While there is much evidence indicating that incorporating a value tilt in the portfolio
enhances the probability for future investment success, such an outcome is not guaranteed.

THE SMALL COMPANY / LARGE COMPANY DIMENSION

As discussed earlier, cumulative returns for small company stocks tend, over longer planning horizons, to
exceed returns for large company stocks. Academic studies suggest that company size (defined as a firms
market capitalizationthe number of outstanding shares multiplied by the price per share) remains an
important determinant of return even after controlling for other variables such as earnings yield and earnings
growth.92 For example, over the period 1973 through 2006, a $1,000 investment in small company stocks
grew to $82,990 versus $31,425 for a comparable investment in large company stocks:

Basu, Sanjoy, The Relationship between Earnings Yield, Market Value and Return for NYSE Common Stocks: Further
Evidence, Journal of Financial Economics (1983), pp. 129-156.

92

50

Figure 3.5: Ending Value of $1,000 from U.S. Large and Small Stocks
19 73 throu gh 20 06

U .S. S m al l C o m pa n y
Stoc ks *
$ 8 2,9 9 0

$ 1 0 0 ,0 0 0

$ 5 0 ,0 0 0
$ 3 0 ,0 0 0
$ 2 0 ,0 0 0

U.S. La rg e Co m p a ny
Sto cks * *

$ 1 0 ,0 0 0

$3 1 ,4 2 5

$ 5 ,0 0 0
$ 3 ,0 0 0
$ 2 ,0 0 0

$ 1 ,0 0 0

$500
$300
$200

$100

197 3

1 976

1 979

19 82

19 85

1 988

19 91

19 94

199 7

200 0

2 003

2 006

* CRSP NY/ A M /NM 6-8

** CRSP NY/ A M /NM 1-2

The graph illustrates that, over the long term, investments in small company stocks have earned a
comparatively high return.93 Nevertheless, the fact that the two lines cross each other in the early 1970s and
draw close to each other during the mid to late 1980s indicates small stocks underperform large stocks over
relatively long periods. Consider, for example, the period 1984 through 1990. S&P 500 (large company)
returns were consistently better than the returns of the 9-10 decile (all stocks with capitalization falling within
the range defined by the smallest 20% of stocks on the New York Stock Exchange). The shaded area indicates
years of small cap stock underperformance relative to the S&P 500:

See, also, Reinganum, Marc R., Portfolio strategies based on market capitalization, The Journal of Portfolio
Management (Winter, 1983), P. 32: One dollar invested in the small firm portfolio at the end of 1962 would have
mushroomed in value to more than $46 at the end of 1980. On the other hand, $1 invested in the large firm portfolio at
the end of 1962 grew in value to slightly more than $4 at the end of 1980. Reinganum attributed the higher returns to
small cap stocks to their greater risk: Over the 18 years of this study, the odds for a small versus large firm doubling in
value were about 10 to 1. On the down side, however, a small firm was almost twice as likely as a large one to experience
a one-year return of 25% or less p. 36. This study was an important precursor to the multifactor models developed in the
mid 1990s which considered the small cap factor to represent a priced risk.
93

51

F ig u re 3.6: La rge v s. Sm a ll Sto ck Re tu rn s

(1 9 8 4 thr oug h 1 9 9 0 )
31.73%

31.69%

24.62%

22.80%
16.61%

18.67%
6.84%

6.27%

10.18%
5.25%

-3.11%

- 6.65%

-9.27%
- 21.57%

19 84

1 985

19 86
S& P 500

1 987

19 88

1989

199 0

Sm all Company

During this period each dollar invested in small company stock94 grew to $1.20, versus $2.60 for each dollar
invested in the S&P 500. Underperformance was both substantial and persistent, with returns for large
company stocks beating small company returns in six out of seven years. This suggests the possibility that the
small stock premium (the difference in performance between small cap and large cap stocks) may be a
spurious artifact of data mining95 (or, may be confounded with other factors). Close examination reveals that
small stocks do not move in tandem with large stocks. Therefore, the benefits of including small stocks in a
portfolio may not flow primarily from an expectation of capturing returns from a priced risk factor, but rather
from the ability to control risk (reduce return variance of the aggregate portfolio).
Some recent research has questioned the existence of a small company premium. Several studies suggest
that the superiority of small stock returns is a myth. The best performing period for small company stocks
(relative to large company stocks) was 1975 through 1983, when the small stock premium (performance of
the smallest 10% of New York Stock Exchange companies minus performance of the largest 10% of New York
Stock Exchange companies) is a huge 21% per year compounded advantage. Some experts argue that if this
period is eliminated from the historical return sequence, the small company premium disappears. The
premium was negative from 1984 through 2003 (small stock returns were , however, strongly positive in
2004),96 although the diversification benefits of adding small company stocks to the portfolio appear to
remain.
94 Small Company stock is all publicly traded securities with market capitalization equal to the bottom two deciles (20%) of
New York Stock Exchange stocks where decile ranking is based on firm size.
95 Black, Fischer, Estimating Expected Return, Financial Analysts Journal (1993), pp. 36-38.
96 Dimson, Elroy, Marsh, Paul, and Staunton, Mike, Low-Cap and Low-Rated Companies, Journal of Portfolio Management
(Summer, 2004), pp. 133-143. The authors of this study contend that the reversal of the size premium extends to most
European stock markets during this period. Robert Arnotts recent study [Arnott, Robert D., Disentangling Size and Value,
Financial Analysts Journal (September/October, 2005), pp. 12-15] points out that the size factor is determined by market
capitalization; but, because market capitalization is the product of sales, earnings or other accounting factor multiplied by
valuation measures like price to sales or price to earnings, the size and value effects are intertwined. For example, a large

52

THE THREE FACTOR MODEL: EMPIRICAL RESULTS

The Fama/French research suggests that investors have the opportunity to capture returns from at least
three risk factors:
The market risk factor - exposure to the risk of equities as opposed to risk-free fixed income
(T-Bills);
The size risk factor - exposure to small company risk as opposed to large company risk; and,
The value risk factor - exposure to value investing as opposed to growth investing.
Investors in the aggregate U.S. stock market as proxied by the S&P 500, from 1964 through 2006, earned
a return in excess of T-Bills of 5.72% per year compounded. However the extra reward (equity risk premium) of
5.72% per year was never precisely realized in any single year. Rather, the investor earned the reward by
tolerating a significant amount of year-to-year volatility. The following graph depicts the year by year
differential returns (Equity minus T-Bill) during the period:
Fig 3.7: Market Risk Premium
1

Market Returns minus Risk Free Returns , 1964-2006

40%

Stoc k Market Outperforms

3 1.8 %

3 1.4%

30%

2 8.1%
25.0 %

24.4 %

27.7%

23.7%

23.1%

21.2%
19.8 %

20%

18 .8%

17.9 %

15.1%

13 .7%

12 .9%

10%

10 .9%

9.9 %

8 .5%

16.4 %

12.3 %
10.5%

9 .7%

8.1%

11.0%

7.1%

5.9 %
4 .2%

1.9%
-0.2 %

0%

- 0.6 %
-2.5%

-2 .6%

-3.6 %

-10%
-11.0%

-12.3%
-14 .8%

-15.1%

-15.0 %
-15.7%

-20%

-19.6 %
-2 1.6%
- 23.8 %

-30%

T-Bills Outperform
-3 4.5%

-40%
1964

1967

1970

1973

1976

1979

1982

1985

1988

1991

1994

1997

2000

2003

2006

1 S&P

50 0
2 3 0 Da yU.S. T re asu ry Bills

There is considerable debate concerning the extent to which investors can expect a significant equity risk
premium in the future. The Equity Risk Premium [ERP] is the expected additional return for making a risky
investment rather than a safe one.97 Current estimates of the ERP range from approximately negative one
percent through positive nine percent, depending on the length of the projection horizon. The debate has farcompany is large either because it has many employees, large sales, valuable assets, etc., or because it is a small
company with a lofty valuation ratio. Typical measures of the size and value effects use a capitalization scale. Arnott,
however, employs an economic scale that measures a firms size relative to its sales, cash flow, dividends, and book value
as a current percentage of the total economy. Arnott concludes: When we separate the size effect from the value-versusgrowth effect, we find that size as measured by market capitalization is far less powerful than is generally believed. And
reciprocally, the value effectbecause some of its efficacy has been siphoned off by the mislabeled size effectis far more
powerful and more consistent than is generally believed.
97 Brealey, Richard A. & Myers, Stewart G., Principles of Corporate Finance. 6th Edition (Irwin McGraw-Hill, 2000), p. 1071.

53

reaching practical consequences for both individual investors, pension plan sponsors, and shapers of public
policy. For example, the debate on creating private Social Security accounts assumes a continued positive
equity risk premium sufficient to generate substantial cost advantages to Social Security participants. An
analysis by the Office of the Chief Actuary of the Social Security Administration produced a geometric ERP
assumption of 4.7% (slightly in excess of the 4.45% compounded ERP from 1964 through 2004). However,
the Social Security Administration invited a review of their projections from three prominent economists. John
Y. Campbell of Harvard produced a forward ERP in the 1.5% to 2.5% per year range; Peter Diamond of MIT and
John Shoven of Stanford both estimated a future compound ERP in the 3% to 3.5% range.98 These estimates
are of great interest because they call into question the so-called superiority of equity investments over fixed
income investment. Given the variance in the year to year ERP evidenced in the chart above, and given the
low estimates of the ERP on a go forward basis, the investor should not be surprised to discover that there may
be long periods in which equity no longer outperforms bonds. Under no circumstances should any individual
investment or any investment strategy or approach be considered a sure thing.
From 1964 through 2006, stocks of small capitalization companies returned 5.16% per year more than
stocks of large capitalization companies. In 29 of the 43 years in this period, small company U.S. stocks
outperformed large company U.S. stocks. The following graph depicts the yearly size-factor premium (return
of small stocks minus return of large stocks).
Fig 3.8: Sm a ll Sto ck Prem iu m
Small Company Stoc k Returns* minus Large C om pany Stock R eturns**, 196 4 through 2 006
50%

Sma ll Sto cks Outp erform

39 .89 %

40%

2 7.8 2%
2 6.9 0%
23 .72 % 2 4. 26%

30%
22 .54 %

2 5. 12%
23 .0 8%

2 0.7 6%
18. 10 %

20%

11 .8 6%
1 0.0 7%

10%

10. 16 %

9.7 1%
8. 25%

7. 93 %

7. 96 %
7. 98%
6 .3 1%

6 .01 %
2.9 6%
0.6 0%

0 .0 4%

0 .68 %

0 .67 %

0 .51 %

1.8 5%
0 .40 %

0%
- 1.8 9%

-2 .83 %
- 5. 03 %
- 5.3 2%

-10%

-7 .96 %

- 8.5 1% - 9.7 1%
- 12 .0 5% - 12 .26 %
- 1 4.6 2%

- 13 .40 %
- 13. 66
-1%
3.8 8%

-20%
- 19 .69 %

-30%

Large Sto cks Outp erform

- 28 .07 %

-40%
1964

1967

1970

1973

1976

1979

1982

1985

1988

1991

1994

1997

2000

2003

2006

* CRSP NY/AM/ NM 6-8

** S&P 500

However, like the market-based equity risk premium, the expected return for investing in small
capitalization stocks is not a sure thing. In 2007, the small stock premium was a negative 5.71%.
Over a comparable period, the premium to the value-style investor (value stock returns growth stock
returns) was 5.43% compounded per year. Although the return to value was greater in 29 out of 43 years in
Derrig, Richard A. & Orr, Elisha D., Equity Risk Premium: Expectations Great and Small, North American Actuarial
Journal (January, 2004).
98

54

the period under evaluation, the years in which value stocks outperformed growth stocks are not identical to
years in which small stocks outperfomed large stocks. Thus, combining small and value factors in the portfolio
may enhance diversification.99
Fig 3.9: V alue Premiu m
Value Stock Returns* m inus Grow th Stoc k Returns* *, 196 4 throug h 2 006
40%

Value Stocks Ou tperform

30%

27.0 8%
2 3.05 %

2 1.4 6%

19 .93 %

18.1 2%

20%

19. 43%
1 6.8 9%

1 6.2 8%

17.116%
7.13 %
1 3.4 3%

1 1.1 0%

9 .06 %

10%

6.08 %

5.8 4%

4.6 8%

14. 41%

13.7 3%
12 .38 %
9. 35%

1 2.2 5%
6 .19 %

8 .78 %

7 .44 %

2.57 %
0. 56%

0 .52 %

0. 27%

0%
-0 .89%
- 2.7 0%

- 3.5 2%
-6. 41%

-10%

- 7.73 % -7. 90%

-1 0.1 9%

- 11.3 9%

- 11 .03 %
-13 .81 %
- 15 .98%

-20%

-1 8.7 5%

-1 9.2 7%

- 18. 41%

Growth Stock s Ou tp erform

-30%

- 29. 65%

-40%
1964

1967

1970

1973

1976

1979

1982

1985

1988

1991

1994

1997

2000

2003

2006

* Fa ma -Fre nc h L arg e Comp a ny Val ue In d ex

** Fama -Fre nc h L arg e Comp a ny Grow th In d ex

The Fama/French three factor model, although subject to the criticism that it is not grounded in economic
theory because it uses stock attributes to explain stock returns,100 has, in the main, provided an attractive
reference point for beginning a discussion about multifactor portfolio construction.101 Current academic
research into the factors driving stock returns remains active. Factor research includes:

macro-economic factors such as interest rates, credit default premiums, inflation and real business
activity;

fundamental factors such as size, value, industry, price/earnings relationships;

technical factors such as the ability of a firms past returns to predict its future returns (a momentum
factor);

the market factor as expressed either in equal weighted or capitalization weighted extensions of the
traditional single factor Capital Asset Pricing Model; and,

statistical factors: mathematical factorizations of returns into (sometimes unobservable) statistical

constructs (that is, linear algebra techniques such as principal components analysis).

Finally, there is an ongoing research effort to relate financial markets with the real economy. To date,
macro-economic models remain generally unsuccessful in explaining the price dynamics of financial assets
The return-to-value in 2007 was a negative 20.53%.
See, for example, Burmeister, Edwin, Roll, Richard & Ross, Stephen, Using Macroeconomic Factors to Control Portfolio
Risk, Working Paper (March 9, 2003): there is no rigorous theory to tell us how traditional accounting variables should
be related to an appropriate measure of risk for computing the risk-return tradeoff. Even if historical empirical
relationships can be uncovered, without the foundation of a rigorous theory one must be concerned that any historical
correlation might be spurious and subject to sudden and material change pp. 1-2.
101 A more detailed discussion of factor models is found on the firms website (www.schultzcollins.com) in the paper
entitled Multifactor Asset Pricing Models and the Rationale for Investing in Value Stocks.
99

100

55

such as stocks and bonds. Often, the models must assume unrealistic values for interest rates, investor risk
aversion, or other critical parameters if they are to calibrate successfully with historically realized financial
returns. This is an important piece of information to remember next time you see an economist or money
manager on TV predicting future stock prices from current economic data (e.g., growth rate in industrial
production, GNP, etc.).102 Recent findings regarding the role of labor income, proprietary (business-owner)
wealth, established consumption (standard of living) levels, and other non-market variables or non-traditional
factors may be of special importance for portfolio design purposes.
The emerging picture suggests that the Fama/French factor model remains a useful structure for
explaining the cross section of stock returns. The list of fundamental factors (market, small & value) has been
extended to include momentum factors as well as bond factors such as the default premium and the termto-maturtiy premium. It appears that at some yet poorly understood level, the factors are strongly interrelated
and may reflect true, but perhaps unobservable, determinants of expected returns. Thus, portfolio
combinations of capitalization weighted and momentum-driven investments like the S&P 500 index with
investments that reflect differing risk characteristics (correlation) such as value, small, and, especially, foreign
investments appear to provide attractive opportunities for investors.
FACTOR LOADING AND UNSYSTEMATIC RISK
It is pertinent at this point to refer back to the earlier discussion of systematic and unsystematic risk, and
to answer the question whether loading a portfolio for, e.g., the value effect does not constitute an increase in
exposure to unsystematic - that is, uncompensated - risk, and is therefore imprudent. The same question
would seem to apply to any deviation from the asset allocation of the current global market. That is, if the
asset allocation of global markets has, by definition, eliminated unsystematic risk, should not every investor
mimic that allocation, and abjure loading for or against any particular risk factor?
The answer would be yes, if every investor had the same risk tolerance and faced an identical set of
economic incentives and objectives. The asset allocation set by world markets represents the most rational
tradeoff of risk and reward only for investors with average risk aversion (the average we speak of here is the
average risk tolerance of all the worlds investors). To the extent that an investor is more or less tolerant of risk
than the average investor, risk loading decisions will differ from those of the global capital markets.103
Furthermore, the asset allocation set by world markets nets against each other all the differing economic
environments faced by each of the worlds investors. Japanese investors, for example, are rewarded differently
for holding U.S. Treasury securities than are U.S. investors. For Americans, Treasury Bills are a risk-free
investment; not so for the Japanese, who must bear currency risk if they own U.S. securities. Similarly, different
investors have different time horizons, face different tax and regulatory regimes, and hold different illiquid
investments (e.g., family real estate, privately owned businesses, etc.).
Cochrane, John H., Financial Markets and the Real Economy, Working Paper, University of Chicago (September 21,
2005), provides a good review of the literature.
103 See, for example, Fama, Eugene F. & French, Kenneth R., Disagreement, Tastes, and Asset Prices, Working Paper,
University of Chicago (February, 2004). The authors argue that equilibrium asset pricing models generate prescriptive rules
for portfolio construction (e.g., build a portfolio that replicates the world capital market) only in the special case where both
informed and uninformed investors hold the set of maximally efficient assets (i.e., all uninformed investors hold fully
indexed portfolios). Problems arise if there are no informed investors or if uninformed investors hold poorly designed
portfolios. When the market share of the informed is small, the bad information of the misinformed can have a big effect
on asset prices. (p. 6). In classical economic theory, the market assures that asset prices reflect true (justified) value
through the mechanism of arbitrage. In practice, however, this means that informed investors must overweight their asset
holdings to compensate for the underweighting of the misinformed. But such an overweighting is not a riskless arbitrage;
and, therefore, the market is probably mostly efficient rather than maximally efficient. If active managers have superior
information, they help to make the market more efficient. This forms a basis for justifying active management but still
leaves open the problem of identifying the active managers with superior skillthat is, which managers will win in the
future.
102

56

One of the critical decisions that an investor must make is whether or not to buy the world market index.
In such a case, the investor may decide not to seek customized financial advice. As one commentator puts it:
To rationalize anything but the market portfolio, you have to be different from the average investor in some
identificable way.104 For each investor, personal preferences and financial positions are integrated in what
economists call the investors utility function. Each investors utility function is, obviously, unique. Through the
process of trading, these unique utility functions are netted against each other and subsumed by the global
capital markets. But the set of global utility functions (if they may be so called) thereby defined differs from
that of any individual. That difference makes it just as easy to argue convincingly that no investor should mimic
the asset allocation defined by world markets, as that every investor should.105
It may be, therefore, rational and prudent to load for such factors as value and size, despite the fact that
such loadings may not be macro-consistent with world markets. The degree to which such loading represents
an increase in exposure to unsystematic risk depends, in large part, on how the loading is effected. To invest
ten percent of a portfolio in a single small foreign firm is indeed to assume a high degree of uncompensated
risk. On the other hand, a ten percent allocation to broadly diversified funds investing in many thousands of
such firms is a different matter.
High standard deviations of a factors returns (based, for example, on mimicking portfolios designed to
replicate the time series of factor returns) suggest that the factor under evaluation:
1. Is priced in the marketplace;
2. Has explanatory value with respect to the return generating process; and,
3. Represents systematic (compensated) as opposed to unsystematic (uncompensated) risk to the
investor.
Factors deemed to be important in explaining stock returns are useful in constructing portfolios. Factor
models imply that stocks with different factor sensitivities will have different average returns. Thus, investors
using a multi-factor approach are, in essence, simply making decisions regarding the risk/return tradeoffs
available in the marketplace.106
As the cost of information continues to drop, new investment products enable investors to load for
increasingly specific risk factors (e.g., European Small Cap Value stocks) without assuming undue
unsystematic risk. Comprehensive diversification across an asset class enables investors to mitigate the
unsystematic risk entailed in factor loading; and, to increase exposure to the specific systematic risk factor(s)
they wish to target.

Cochrane, John H., Portfolio advice for a multifactor world, Economic Perspectives Federal Reserve Bank of Chicago
(3rd Quarter, 1999), p. 73.
105 This point is made by Harry Markowitz in a recent article: if the universe consists of, say 10,000 securities, then if all
securities are to be demanded by someone, this universal efficient frontier must contain at least 10,000 segments. If
investors have sufficiently diverse risk tolerances, they will choose portfolios on different segments. Some will prefer
portfolios on one or another of the typically less diversified high-risk/high-return segments. Others will select portfolios on
one or another of the typically more diversified lower-risk segments. The market is an average, weighted by investor
wealth, of portfolios selected from these diverse segments. Although it is mathematically possible for this average to
accidentally fall on the efficient frontier, such an outcome is extremely unlikely. Markowitz, Harry, Market Efficiency: A
Theoretical Distinction and So What? Financial Analysts Journal (September/October, 2005), p-. 27. There is a long
history of debate on the issue of whether broad market proxies such as the S&P 500 are themselves efficient portfolios.
See, for example, Benninga, Simon, Financial Modeling (MIT Press, 1997), pp. 117-123.
106 Focardi, Sergio M. & Fabozzi, Frank J., The Mathematics of Financial Modeling & Investment Management (Wiley
Finance, 2004), p. 539.
104

57

C H A P T E R FO U R : B U I L D I N G T H E P O R T FO L I O

The asset allocation decision forms the foundation of the portfolio.107 How should the investor make the
asset allocation decision?
Effective diversification makes possible the measurement and control of overall portfolio risk. On the
assumption the investor intends to build a diversified portfolio, the portfolios asset allocation structure flows
from a decision about how much portfolio risk the investor is willing to bear. Strategic asset allocation is a
function both of the investors required return and the investors tolerance for risk. Trivially, an allocation with
an expected return below that required to achieve financial goals is inappropriate. It is easy to see this type of
mistake. Equally insidious, however, are allocations that produce more risk than necessary in light of the
purposes, goals and economic circumstances of the portfolio. The greater the risk, the greater is the
uncertainty of future results. That is to say, the portfolios asset allocation policy is best tailored to meet the
investors return objectives without generating a potentially dangerous risk gap, where risk gap is defined as
the voluntary election to assume risk greater than necessary. More money is better than less; however, the
pursuit of more money may be dangerous to your wealth. Effective portfolio management requires both
measurement and management of risk.
One way for an investor to find an asset allocation suitable to his risk tolerance is to track the behavior of
a set of model portfolios. The investor considers the historical returns achieved by various diversified
portfolios each each of which differs according to their percentage allocation to fixed income (bonds) versus
equities (stocks). A low variance portfolio owns a greater percentage of fixed income investments, while a high
variance portfolio allocates more to stocks. Using this method, the investor can test drive the consequences
of an asset allocation election along a continuum of risk/return choices.
THE RISK / RETURN CONTINUUM
The historical performance of several sample portfolios, based on quarterly returns series from 1973
through 2006,108 is shown in the tables and graph below. The first table illustrates the distribution of each
See, for example, Farrell, James L., Portfolio Management: Theory and Application 2nd Edition (Irwin McGraw-Hill, 1997),
p. 272: Determining the asset mix that best suits the risk-return objective of the investor is the most important decision in
meeting the longer-range goals of the investment plan. Determining the asset mix without reference to the investors
liabilities (including consumption targets) is, however, comparable to making financial decisions by looking only at the left
hand side of a balance sheet. Liabilities should be explicitly stated and well defined so that the investor understands the
portfolios required return. A goal expressed in vague terms such as growth, or making a lot of money, or beating the
S&P 500, is not sufficient to allow for the construction of a portfolio that has a reasonable probability of prudently
matching assets to liabilities. This is an important but complex extension of the topics in this introductory monograph. For
further information, see the article in the Investment Quarterly Volume 11, Issue 2 (2005) Planning for Retirement
Income, found at the SCLC website (www. schultzcollins.com/files/IQ2005Q2.pdf); and Howard, Ronald & Lax Yoel,
Strategic Asset Allocation in the Presence of Uncertain Liabilities, Modern Investment Management (John Wiley & Sons,
2003), pp. 110-135.
108 Data Sources:
107

Asset Class
S&P 500 Index
One Year Fixed
Five Year Fixed
International
Small Cap Stocks
U.S. 9-10 Index
U.S. Small Cap
Value Stocks
U.S. Large Cap
Value Stocks

Return Series
Results as reported by Ibbotson & Sinquefield: Stocks, Bonds, Bills and Inflation.
One Year Constant Maturity US T-Bills, as reported by Ibbotson Associates.
Lehman Brothers Intermediate Term Govt/Credit, as reported by Ibbotson Associates.
DFA International Small Company Index, as reported by Dimensional Fund Advisors.
Center for Research in Security Prices (CRSP) 9-10 Index, as reported by Ibbotson Associates.
Fama - French Small Cap Value, as reported by Ibbotson Associates.
Fama - French Large Cap Value, as reported by Ibbotson Associates.

58

portfolio among various asset classes.109 The portfolio labels reflect their ratio of equity to fixed income
investments (that is, the 60/40 portfolio is 60% invested in equities, and 40% invested in fixed income
securities). The table describes how stocks and bonds are distributed into specific asset classes. For example,
the 60/40 portfolio invests 10% each in the S&P 500, U.S. Large Cap Value Stocks, U.S. Small Cap Stocks,
U.S. Small Cap Value Stocks, International Large Cap Stocks, and International Small Cap Stocks. On the Fixed
Income side, it invests 30% in One Year Fixed Income securities and 10% in Five Year Fixed Income securities.
Taken together, the six model portfolios define a continuum of allocations from 100% fixed income to 100%
equity.

Figure 4.1: Distribution of Model Portfolios Among Asset Classes

Asset Class
S&P 500 Stock Index
U.S. Large Cap Value Stocks
Decile 9-10 U.S. Small Cap Stocks
U.S. Small Cap Value Stocks
Int'l Large Cap Stock
Int'l Small Cap Stock
One Year Fixed Income
Five Year Fixed Income

0/100
0%
0%
0%
0%
0%
0%
100%
0%

Portfolio Equity/Fixed Income

20/80 40/60 60/40 80/20
4%
6.5% 10% 12.5%
4%
6.5% 10% 12.5%
4%
6.5% 10% 12.5%
4%
6.5% 10% 12.5%
2%
7%
10%
15%
2%
7%
10%
15%
60%
50%
30%
15%
20%
10%
10%
5%

100/0
17.5%
16.5%
16.5%
16.5%
16.5%
16.5%
0%
0%

The next table illustrates results realized over the thirty-four year period. Annualized Return measures the
mean annual compound rate of return. Standard Deviation measures the degree of investment risk realized by
each portfolio. Ending Value of $1,000 describes the accumulation of $1,000 invested in each portfolio on
January 1, 1973, as of December 31, 2006. The divergence of returns and cumulative results over the thirtyfour year period is dramatic:

Figure 4.2: Historical Performance of Model Portfolios, 1973 - 2006

Portfolio Equity/Fixed Income
Annualized Return
Standard Deviation
Ending Value of $1,000

0/100
6.95%
3.34%
$9,807

20/80
8.94%
4.72%
$18,406

40/60
10.45%
7.56%
$29,370

60/40
11.97%
10.97%
$46,687

80/20
13.29%
14.52%
$69,500

100/0
14.49%
17.96%
$99,597

The $99,597 earned by the 100% equity portfolio for each dollar invested on 1/1/73 is an attractive
number indeed. However, investors should consider how disconcerting it would be to experience the worst
annual losses that the all-equity portfolio generated over the period. The Arab Oil Embargo years of 1973 and
1974 were particularly distressing.

Returns are proxied by indexes; and, therefore, should not be construed as a track record of any actual portfolio.
Indexed investments were not available to investors throughout the entire period. The annual return and standard deviation
calculations are based on monthly returns for portfolios that were rebalanced quarterly during each one-year holding
period.

109

59

Figure 4.3: Annual Portfolio Performance

Annual Portfolio Performance: 34 Calendar Years 1/1/73 - 12/31/06
Year

0/100

20/80

40/60

60/40

80/20

100/0

1973

7.45%

1.39%

-3.08%

-8.21%

-12.88%

-17.93%

1974

8.55%

1.50%

-4.85%

-11.40%

-17.88%

-24.27%

1975

7.08%

16.33%

24.30%

32.88%

40.97%

49.56%

1976

6.22%

13.64%

17.00%

22.45%

26.12%

32.17%

1977

6.11%

7.61%

12.23%

15.05%

19.53%

21.19%

1978

8.43%

9.43%

14.59%

17.60%

22.44%

24.59%

1979

11.15%

13.03%

14.72%

17.06%

18.50%

21.82%

1980

12.60%

14.21%

17.91%

20.84%

24.23%

27.41%

1981

16.23%

12.97%

10.96%

8.69%

6.39%

4.52%

1982

13.51%

18.04%

17.00%

18.41%

18.00%

19.60%
31.01%

1983

9.98%

13.96%

18.28%

22.51%

26.76%

1984

11.63%

11.35%

10.42%

9.65%

8.85%

7.85%

1985

8.98%

16.63%

23.37%

29.81%

36.99%

42.08%

1986

6.86%

11.87%

17.93%

22.67%

29.05%

32.03%

1987

6.90%

5.73%

7.66%

7.85%

9.62%

8.98%

1988

7.81%

11.07%

14.79%

18.30%

21.97%

25.38%

1989

9.06%

12.66%

14.96%

17.76%

20.25%

23.03%

1990

8.29%

3.63%

-1.64%

-6.59%

-11.82%

-16.53%

1991

6.29%

12.95%

15.45%

19.89%

22.81%

27.60%

1992

4.03%

6.54%

5.88%

7.02%

6.52%

8.48%

1993

3.51%

8.28%

12.35%

16.37%

20.71%

23.94%

1994

5.16%

3.07%

3.88%

3.47%

3.92%

3.51%

1995

6.31%

12.42%

14.28%

18.07%

20.38%

24.55%

1996

5.67%

7.43%

8.88%

10.70%

12.06%

14.41%

1997

5.80%

9.98%

11.14%

14.11%

15.37%

19.44%

1998

5.28%

6.91%

7.93%

9.04%

10.22%

11.20%

1999

5.10%

6.72%

10.68%

13.55%

17.27%

19.97%
-6.77%

2000

6.34%

4.71%

1.42%

-1.13%

-4.24%

2001

3.84%

4.94%

2.71%

2.19%

0.22%

0.11%

2002

2.09%

-0.72%

-5.09%

-9.19%

-13.18%

-18.04%

2003

1.26%

11.23%

20.43%

29.84%

39.19%

48.18%

2004

1.79%

5.85%

9.99%

13.96%

18.16%

21.75%

2005

3.54%

4.21%

6.06%

7.31%

9.06%

10.10%

2006

5.02%

7.93%

11.55%

14.79%

18.37%

21.42%

60

To be truly successful, portfolios must often generate returns over and above the cost of living. Adjusting
nominal annual returns for inflation provides a useful picture of portfolio results:

Figure 4.4: Annual Constant Dollar Portfolio Performance

Annual Portfolio Performance Adjusted for Inflation: 34 Calendar Years 1/1/73 - 12/31/06
Year

0/100

20/80

40/60

60/40

80/20

100/0

1973

-1.24%

-6.81%

-10.92%

-15.63%

-19.93%

-24.56%

1974

-3.25%

-9.54%

-15.20%

-21.03%

-26.81%

-32.51%

1975

0.06%

8.71%

16.15%

24.17%

31.73%

39.76%

1976

1.34%

8.42%

11.63%

16.83%

20.33%

26.10%

1977

-0.62%

0.79%

5.11%

7.76%

11.95%

13.51%

1978

-0.55%

0.37%

5.10%

7.87%

12.31%

14.27%

1979

-1.90%

-0.25%

1.25%

3.31%

4.58%

7.51%

1980

0.18%

1.61%

4.91%

7.51%

10.52%

13.36%

1981

6.69%

3.70%

1.85%

-0.23%

-2.34%

-4.06%

1982

9.28%

13.64%

12.64%

14.00%

13.60%

15.15%

1983

5.95%

9.79%

13.95%

18.03%

22.13%

26.21%

1984

7.39%

7.12%

6.22%

5.48%

4.71%

3.75%

1985

5.02%

12.40%

18.88%

25.09%

32.01%

36.92%

1986

5.66%

10.62%

16.62%

21.30%

27.61%

30.56%

1987

2.39%

1.26%

3.12%

3.29%

4.99%

4.38%

1988

3.25%

6.37%

9.93%

13.29%

16.81%

20.07%

1989

4.22%

7.66%

9.86%

12.53%

14.91%

17.57%

1990

2.06%

-2.33%

-7.30%

-11.97%

-16.90%

-21.33%

1991

3.13%

9.59%

12.02%

16.33%

19.16%

23.81%

1992

1.10%

3.54%

2.90%

4.00%

3.52%

5.42%

1993

0.74%

5.39%

9.35%

13.26%

17.48%

20.62%

1994

2.42%

0.39%

1.17%

0.78%

1.22%

0.81%

1995

3.68%

9.64%

11.45%

15.15%

17.40%

21.46%

1996

2.27%

3.98%

5.38%

7.14%

8.46%

10.73%

1997

4.03%

8.14%

9.28%

12.20%

13.44%

17.44%

1998

3.61%

5.21%

6.22%

7.31%

8.47%

9.44%

1999

2.36%

3.93%

7.78%

10.58%

14.20%

16.84%

2000

2.85%

1.28%

-1.90%

-4.37%

-7.37%

-9.82%

2001

2.25%

3.33%

1.14%

0.63%

-1.31%

-1.42%

2002

-0.28%

-3.02%

-7.29%

-11.30%

-15.19%

-19.94%

2003

-0.60%

9.18%

18.21%

27.44%

36.62%

45.45%

2004

-1.42%

2.52%

6.53%

10.36%

14.43%

17.91%

2005

0.12%

0.76%

2.55%

3.77%

5.46%

6.46%

2006

2.41%

5.25%

8.79%

11.95%

15.44%

18.41%

61

The constant dollar adjustment shows that inflation is a threat to the portfolio that is both subtler and
more persistent than sudden shocks to the domestic or global financial markets. Inflation strikes hardest at
the 100% fixed income portfolio. This low variance portfolio traditionally appeals to the most risk-averse
investors.110 However, over the long term, the safest portfolio is perhaps the most risky. For example, when
portfolio returns are measured over five-year periods, the 100% fixed income portfolio experienced the
greatest number of real losses:

Figure 4.5: Overlapping Five-Year Period Constant Dollar Portfolio

Performance
30 Overlapping Five Year Periods 1/1//73 - 12/31/06
Starting In

0/100

20/80

40/60

60/40

80/20

100/0

1973

-0.75%

0.03%

0.58%

0.82%

0.79%

0.37%

1974

-0.62%

1.52%

3.97%

5.89%

7.84%

9.06%

1975

-0.34%

3.53%

7.72%

11.74%

15.82%

19.70%

1976

-0.32%

2.14%

5.55%

8.57%

11.83%

14.79%

1977

0.72%

1.23%

3.63%

5.19%

7.25%

8.69%

1978

2.65%

3.69%

5.07%

6.38%

7.57%

9.00%

1979

3.96%

5.57%

6.79%

8.32%

9.39%

11.19%

1980

5.85%

7.09%

7.81%

8.77%

9.41%

10.40%

1981

6.86%

9.27%

10.55%

12.11%

13.37%

14.65%

1982

6.65%

10.69%

13.58%

16.58%

19.60%

21.93%

1983

5.27%

8.17%

11.59%

14.30%

17.73%

19.56%

1984

4.73%

7.48%

10.79%

13.37%

16.69%

18.38%

1985

4.10%

7.59%

11.54%

14.85%

18.88%

21.37%

1986

3.51%

4.61%

6.13%

7.06%

8.37%

8.64%

1987

3.01%

4.42%

5.28%

6.16%

6.89%

7.49%

1988

2.74%

4.88%

5.23%

6.31%

6.59%

7.71%

1989

2.24%

4.69%

5.12%

6.30%

6.71%

7.81%

1990

1.88%

3.23%

3.40%

3.98%

4.04%

4.54%

1991

2.20%

5.65%

7.28%

9.72%

11.48%

14.03%

1992

2.03%

4.54%

5.98%

7.93%

9.40%

11.51%

1993

2.62%

5.46%

7.26%

9.58%

11.42%

13.94%

1994

3.20%

5.42%

6.64%

8.40%

9.66%

11.75%

1995

3.19%

6.16%

8.00%

10.43%

12.34%

15.09%

1996

3.02%

4.49%

5.28%

6.41%

7.14%

8.44%

1997

3.02%

4.36%

4.42%

5.08%

5.14%

5.95%

1998

2.15%

2.11%

1.04%

0.26%

-0.81%

-1.87%

1999

1.31%

2.86%

3.23%

3.76%

3.88%

3.88%

2000

0.55%

2.58%

2.98%

3.72%

3.92%

4.07%

2001

0.01%

2.48%

3.90%

5.43%

6.65%

7.58%

2002

0.04%

2.86%

5.43%

7.70%

10.05%

11.60%

The salient point is that, in spite of the fact that the 1970s was arguably the worst decade for the U.S.
stock market since the Great Depression, balanced portfolios did well over this period, and over most
subsequent five year periods. Poorly diversified portfolios, tilted strongly towards either 100% fixed income or
100% equity, however, did not. When the bear market of 2001-2002 is factored in, the advantage of prudent,
For a more detailed discussion of the combined effects on portfolios of inflation, trading costs and taxes, see TAXES,
INFLATION AND TURNOVER below.
110

62

balanced and globally diversified portfolios is fully borne out.111 All-equity portfolios appeal to investors
wishing to accumulate wealth quickly. Ironically, however, the return variance properties of an all-equity
portfolio require the investor to possess both great patience and extraordinary risk tolerance.
As the graph below makes manifest, risk and return are quite tightly related (note that the y-axis is
logarithmic in scale). Increasing portfolio exposure to equities increases the period-to-period amplitude of
changes in portfolio value. The $99,597 terminal value was not a free lunch.

Fig 4.6: Ending Value of $1,000

100/0
$99,597

1973 th rough 2006

$100,000

80/20
$69,500

0 /100
$50,000

$30,000

2 0/80

60/ 40

4 0/60

$46, 687
40/60
$29,370

6 0/40
8 0/20

$20,000

20/80
$18,406

1 00/0
$10,000

0/ 100
$9,807

$5,000
$3,000
$2,000

$1,000

$500

$300
$200

$100

19 73

197 6

19 79

198 2

19 85

19 88

19 91

19 94

199 7

2000

200 3

20 06

DEFINING ASSET CLASS WEIGHTINGS

The Risk/Return Continuum is useful as a heuristic tool, either to gauge risk tolerance or to convey the
cost in volatility that must be paid to achieve a required return. Once the investor determines the preferred
risk and return characteristics of the portfolio, the job of portfolio building and implementation begins.
The first step is the selection of asset classes. Numerous asset classes are absent from the model
portfolios of the risk/return continuum. These include Real Estate, Emerging Markets, and International
Bonds. Most investors will want to include one or more of these asset classes for the sake of diversification.
As a general rule, expanding the opportunity set of investments will enhance the ability to customize a portfolio
that can produce a satisfactory future outcome.

One cannot conclude that the 100% equity portfolio, which generated the highest terminal wealth in the absence of
cash flows, would also generate the greatest amount of ending wealth in the presence of periodic distributions. The
presence of cash flow requirements may alter the optimal asset allocation because each cash flow acts as a multiplier on
downside returns and as a cap on upside returns. The Equity is better because it outperforms Bonds argument is a
potentially dangerous approach to wealth management in the presence of liquidity needs.

111

63

On the other hand, many investors have unique preferences or constraints that limit asset allocation
policy. For example, the investor may believe that U.S. investments are intrinsically superior to foreign
investments; or that investing in foreign companies facilitates the transfer of domestic job opportunities to
overseas competitors. Alternately, some investors are leery of certain U.S. Large Company stocks, and may
wish to avoid investing in companies engaged in businesses or practices they consider morally reprehensible
or environmentally destructive. Although engaged in many, varied and far-flung businesses, for example,
General Electric is active in defense contracting and nuclear energy. Similarly, RJR/Nabisco is both a food and
beverage as well as a tobacco company. Either firm might be objectionable to socially conscious investors.
The asset allocation decision increases in complexity when an investor owns illiquid financial assets.
Illiquid financial assets include residential and commercial real estate, closely-held business interests,
annuities, long-term certificates of deposit, and limited partnership units. Alternatively, even when financial
assets are readily tradeable, the capital gains tax on sales of highly appreciated positions in stocks or mutual
funds may make sales impracticable. All these factors affect which asset classes a portfolio should own, as
well as each assets proportionate weighting within the portfolio.112
Once the investor identifies and selects appropriate asset classes, he must determine suitable target
weightings. Adoption of one of the nave allocations used in the Continuum Model Portfolios may be
inappropriate. At this point, the advantages of portfolio factor loading come into play. By customizing
exposures (value vs. growth; small vs. large, foreign vs. domestic, etc.), investors can fine tune their portfolio
so that it can better support cash flows, wealth accumulation, or other objectives. For example, some
investors may have a dynastic or multi-generational planning horizon for family wealth accumulation. They
may wish to invest solely in U.S. stocks because of their long-term high expected returns; but they are not
prepared to assume the amount of risk exhibited in a 100% U.S. large company portfolio. By diversifying the
portfolio by including U.S. small company and value risk factors, they can fine tune the portfolios asset
allocation so that it is better matched to their objectives and preferences.
Factor loading, if it is not to expose the portfolio to substantial unsystematic risk, should be both careful
and sophisticated. Indeed, intelligent factor loading may enable investors selecting a fixed income percentage
allocation to increase the expected return to the equity portion of their portfolios. As one commentator puts it:
Risks can be productive if they are expected to generate return, or unproductive when they
are too large or unintended. Thus, knowing the level of risk in a portfolio is not enough. The
investor needs to measure where the risk is coming from.113

This introductory monograph does not consider the hedging aspects of investment decision making. For example, a
college professor or civil service employee may have labor income similar to a bond. An investor with this type of labor
wealth may have a high demand to hold equities in the investment portfolio. An entrepreneur, however, may have an
income stream with equity-like characteristics and, therefore, may have a high demand to hold fixed income instruments in
the portfolio. Hedging objectives can be critically important in the design of the investment program; and picking the
money managers with the highest realized returns to date in the hope that positive returns can also hedge shocks to labor
income may be counterproductive to both consumption and wealth accumulation objectives.
113 Litterman, Bob Risk Measurement, Modern Investment Management (John Wiley & Sons, 2003), pp.33-34.
112

64

C H A P T E R F I V E : I N V E ST M E N T ST R AT E G I E S A N D I N V E ST M E N T V E H I C L E S

Portfolios are aggregations of specific investments. Once an investor determines the portfolios asset
allocation, he must select investments to meet the asset class weighting targets. An important decision is
whether to purchase individual securities (that is, individual stocks, bonds, options contracts, etc.) or pooled
investment vehicles (that is, mutual funds, exchange-traded funds, etc.). For all but the largest portfolios,
pooled investment vehicles are preferable if the investor wishes to achieve a reasonable level of
diversification.114
Pooled investment vehicles such as mutual funds offer efficient diversification for each dollar invested.
Even small dollar amounts invested in a broadly diversified fund can, in effect, spread investment risks across
many individual issues within an asset class. A U.S. investor attempting to purchase a broad sample of Pacific
Rim small company stocks, for example, would face daunting information and trading costs if he elects to
purchase individual securities. Mutual funds, however, offer significant economies of scale because they
spread costs across thousands of customers, and risk over hundreds of securities.
When selecting a mutual fund, however, it is important to monitor how closely the funds performance
mirrors the performance of the asset class it is intended to represent within the portfolio. This is not a simple
matter of consulting the funds prospectus or its marketing materials. The behavior of funds, and thus their
reliability as a representative of a given asset class, can change significantly for a number of reasons:
Fund investment objectives, policies, and portfolio holdings can change over the course of a
market cycle;
Fund investment objectives, policies, and portfolio holdings can change with a transition to a
new manager, or with a change in the funds analytical support team;
Funds may experience style drift as the corporate capitalization and accounting ratios of
securities change over time;
Funds define asset class boundaries differently. For example, there is disagreement within
the financial community about where to draw the line between small cap and mid cap
companies.
Prospectuses often give fund management wide latitude to purchase different kinds of
securities. Many funds, for example, invest in international equities or bonds despite the fact
that their marketing literature suggests that they are domestic funds.
To reach a judgment about how well a funds returns track those of a comparable asset class, the investor
must use statistical methods. A commonly used evaluation metric is the coefficient of determination, or R2
statistic.115 R2 is a measure of how closely the variations in return of one data series - e.g., an index - explain
variations in a second data series - e.g., a mutual fund. An R2 statistical value close to 100 evidences the
success of the fund in capturing the returns of the asset class it represents. Further, the R2 statistic provides
data on the degree to which investment vehicles chosen to meet asset class weighting targets have strayed
114 Consider, for example, the research findings of Burton Malkiel: to get to where idiosyncratic risk asymptotically
touches the systematic risk lineyou need about 10 times as many stocks as before, or 200 stocks. If you want to get the
returns of the asset class itself, of course, indexing would be the best strategy. Malkiel, Burton G., How Much
Diversification is Enough? Equity Portfolio Construction (AIMR, 2002), p. 26.
115 The square root of R2 is the correlation statistic. The reader should note that a high correlation to a benchmark,
although of great importance, may not be as critical as the selection of the benchmark used as a proxy for the asset class.
For example, a small cap fund emulating the Citibank S&P Small Company Value index will perform very differently from a
fund emulating the Russell 2000 small company index despite the fact that they are both small company index funds.
Choice of appropriate benchmarks is a critical decision for asset allocation policy, as well as for investment manager
selection and retention policy. These issues, however, lie outside the scope of this discussion.

65

from their intended purposes - a state of affairs which, left uncorrected, might skew the portfolios future asset
allocation.

ACTIVE VERSUS PASSIVE MANAGEMENT

Investment management styles fall into one of two basic categories--active and passive:
Active management attempts to achieve superior returns by identifying mispriced securities.
Superior returns follow from a willingness to disagree with market prices, concentrating
holdings in a limited number of securities. The manager embraces asset concentration risk
believing that, over time, other market participants will identify the mispricing, and drive the
securitys price to its true price, generating better than market returns for the fund.
Passive management attempts to earn market returns by buying all (or a statistically
representative sample) of the securities within an asset class or index. Generally, passive
management does not try to beat the market; but, rather seeks to match the returns of the
market.
The distinction between active and passive investment strategies is nicely captured by the following
distinctions: passive investment management consists of tracking the market, without attempting to
anticipate its evolution.the objective of active investment management is to perform better than the market,
or better than a benchmark that is chosen as a reference.116

ACTIVE MANAGEMENT
The fund prospectus defines the investment objective of an actively managed fund. Actively managed
funds perform significant research and security analysis, and tend to experience relatively high portfolio
turnover. A limited number of securities are selected from the available investment universe (the opportunity
set). Selection and timing of security purchases rely on either fundamental analysis (business cycle, inflation
and interest rate forecasting), or on technical analysis (forecasting stock market trends based on price and
volume movements). The funds management attempts to beat the market.117 Purchase and sale of individual
securities is based on macroeconomic capital market forecasts as well as forecasted inputs to pricing and
asset volatility models. Success or failure of investment decisions is largely a function of the accuracy of
analysts predictions and on the integrity of their security valuation models.118
As noted earlier, under the Prudent Investor Rule the investor must justify any risks and costs attributable
to active management:
Active strategiesentail investigation and analysis expenses and tend to increase general
transaction costs, including capital gains taxation. Additional risks also may result from the
Amenc, Noel & Le Sourd Veronique, Portfolio Theory and Performance Analysis (John Wiley & Sons, 2003), pp. 6-8.
Active managers seek to add value by beating their comparative benchmarks. Investors, however, should evaluate their
personal objectives prior to portfolio design. Is the objective to beat a market or is the objective to solve an intertemporal
cash flow problem or a wealth accumulation objective?
118 A vast amount of literature explores the accuracy of analyst forecasts, the extent to which forecasts may be biased, and
the impact of the recent SEC disclosure obligations (Regulation FDs requirement for corporations to disclose material
information publicly and uniformly). Dreman, David N. & Berry, Michael A, Analyst Forecasting Errors and Their
Implications for Security Analysis, Financial Analysts Journal (May/June, 1995), pp. 30-41 provides a good historical
survey of the research. A comprehensive update is found in Francis, Jennifer, Chen, Qi, Willis, Richard H. & Philbrick,
Donna R., Security Analyst Independence (Research Foundation of CFA Institute, 2004). The investor wishing to select
active managers should be sufficiently skilled so that he can, at a reasonable confidence level, identify managers with
positive forecasting abilities.
116
117

66

difficult judgments that may be involved and from the possible acceptance of a relatively high
degree of diversifiable risk. If the extra costs and risks of an investment program are
substantial, these added costs and risks must be justified by realistically evaluated return
expectations. Accordingly, a decision to proceed with such a program involves judgments by
the trustee that:
Gains from the course of action in question can reasonably be expected to compensate for its
additional costs and risks;
The course of action to be undertaken is reasonable in terms of its economic rationale and its
role within the trust portfolio; and
There is a credible basis for concluding that the trustee - or manager of a particular activity possesses or has access to the competence necessary to carry out the program and, when
delegation is involved, that its terms and supervision are appropriate.119
Until recently, investment orthodoxy focused on two methods for controlling risk and enhancing portfolio
returns. Traditional wisdom holds that successful portfolios derive either from superior security selection, or
astute market timing decisions, or both. A preponderance of academic evidence, however, indicates that it is
difficult to sustain consistent market beating performance by using either of these approaches.

Fundamental Analysis
The Efficient Market Hypothesis represents a challenge to classical approaches to stock investing.120
Financial analysts had long used fundamental analysis to locate mispriced securities. At the heart of such
systems lies the belief that analysts can use accounting ratios, valuation techniques, and data from corporate
financial statements and economic reports to develop algorithms for discovering overvalued or undervalued
stocks. Good fundamental analysis entails investigating a firms financial statements, industry position,
general economic trends, competitive advantages, earnings prospects, etc., to forecast its future profitability.
Historically, many believed that financial analysts were best positioned to predict future stock price
movements. Their in-depth knowledge and close observations of particular firms or industries justified trading
(buy/sell) recommendations. Traditionally, the analyst would alert investors - through their brokers - to
undiscovered buying opportunities or unsuspected problems.121
The widespread use of fundamental analysis by the major institutions that dominate trading is an
important reason why bargains can no longer easily be found. The more numerous and more skillful analysts
become, the more difficult it is to earn abnormal profits. As one Wall Street observer comments:

Restatement of the Law, Third, of Trusts (Prudent Investor Rule), p. 30.

For example, Benjamin Graham, a prominent advocate of techniques of fundamental analysis, writes, in 1976: I am no
longer an advocate of elaborate techniques of security analysis in order to find superior value opportunities. In the old
days any well-trained security analyst could do a good professional job of selecting undervalued issues through detailed
studies; but in the light of the enormous amount of research now being carried on, I doubt whether in most cases such
extensive efforts will generate sufficiently superior selections to justify their cost. To that very limited extent Im on the side
of the efficient market school of thought now generally accepted by the professors. A Conversation with Benjamin
Graham Financial Analysts Journal (September/October, 1976), p. 22.
121 It was not that long ago that investors valued brokers because of their timely tips. Money managers were selected for
their ability to act on insider information gleaned from their positions on boards of directors as well as from their
relationships with corporate management. Today, clients who trade on material nonpublic information provided by their
brokers end up like Martha Stewart. In many respects the role of the broker has changed from a vigilant watchman of
market developments who seeks to protect investors from adversity, or who alerts clients to emerging opportunity, to
something akin to a waiter announcing the house specials on the current investment menu.
119
120

67

...the movement of increasing amounts of money into professional management...would

make it just that much more difficult for us to capture rewards for our clients pocketbooks.
With competition for information becoming ever more intense, professional managers were
destined to have a hard time in trying to outperform one another. We could not beat the
market because we were rapidly becoming the market.122
Academic evidence raises questions whether analysts can add value for retail investors in a market
dominated by institutional analysts:
Discovery of good firms does an investor no good in and of itself if the rest of the market also
knows those firms are good. If the knowledge is already public, the investor will be forced to
pay a high price for those firms and will not realize a superior rate of return. This is why
fundamental analysis is difficult. It is not enough to do a good analysis of a firm; you can
make money only if your analysis is better than that of your competitors because the market
price is expected already to reflect all commonly available information.123
The existence of large numbers of savvy analysts who ensure that stock prices correctly and
instantaneously reflect available information creates an investment environment in which successful security
selection is extremely difficult.

Technical Analysis and Market Timing Systems

Technical analysts differ from financial analysts in that they look primarily at information about a
companys stock trading patterns, such as price, volume and other market-related trends. These analysts
generally ignore accounting information and macro-economic data. However, the central question is whether a
careful analysis of past stock price movements and overall market trends can create a successful investment
strategy. 124
In a pioneering 1953 statistical study, Maurice Kendall considered this question. Kendall found no
evidence that any statistically meaningful patterns could be found in stock prices. Securities prices trace a
random walk. On any given day, it is equally likely that the price of a stock would increase or decrease, no
matter what the stocks recent performance.125 Such randomness in price changes is characteristic of an
Bernstein, Peter. Capital Ideas: The Improbable Origins of Modern Wall Street. Maxwell Macmillan, New York (1992), p.
140.
123 Bodie, Zvi, Kane, Alex, and Marcus, Alan J. Investments. Irwin, Burr Ridge, Illinois (1993), p.364.
124 Technical analysis underlies two strategies: (1) market timing; and, (2) trade execution. Market timing refers to tactical
asset allocation strategies such as sector rotation, fixed income duration management in light of interest rate forecasts,
equity beta management in light of macro economic forecasts, and so forth. Tactical asset allocation assumes a
willingness to deviate from the portfolios long-term strategic asset allocation. The extent of the deviation is a function of
the investors confidence in his or her forecasting ability. Unlike fundamental analysts, technical analysts justify such
deviation based on market-related information rather than on exogenous accounting / macro-economic information. This
essay does not discuss the use of technical analysis in the formation of portfolio trade execution strategies. There is little
question that technical analysis has great value in the areas of portfolio implementation and trading strategies. See, for
example, Hasbrouck, Joel, Empirical Market Microstructure, Oxford University Press (2007), p. 4: At a single instant there
may be many prices, depending on direction (buying or selling), the speed with which the trade must be accomplished, the
agents identity or other attribute, and the agents relationship to the counterparty. Close attention to price volatility,
volume, evidence of buy/sell imbalances in the market, and so forth, may be critical to implementing successful trading
strategies. See, also, Schwartz, Robert A. & Francioni, Reto, Equity Markets in Action: The Fundamentals of Liquidity,
Market Structure & Trading (John Wiley & Sons, 2004), pp. 95-96.
122

Kendall, Maurice. The Analysis of Time Series, Part I: Prices., Journal of the Royal Statistical Society, London. Vol. 96
(1953), pp.11-25. Subsequent research indicates that long-term price changes evidence some degree of predictability.
See, for example, Campbell, John Y. & Shiller, Robert J., The Dividend-Price Ratio and Expectations of Future Dividends
125

68

efficient, rational market in which current prices reflect all information. According to this line of research, the
markets price movements are not irrational responses to events; rather, it is the events themselves that are
unpredictable. In such a market, only new information can alter investor perceptions and, by definition, new
information arises in an unpredictable fashion.
Some Technical stock analysts promote market timing systems. Market timing is the attempt to align
portfolio exposure to market risk factors in anticipation of predicted changes in security prices. The lure of
market timing is strong. It promises a system that generates gains and avoids losses. Market timing
vocabulary is pervasive. It is found in many articles written by the popular press, and is constantly broadcast
over radio and TV programs.
Claims of market timing ability are a fruitful area for independent, third party investigation because
standard statistical tests can readily validate or invalidate such claims. In general, market timers justify their
asset management strategies by advancing three assertions:
1. The decision maker or advisor, in fact, possesses market timing ability;
2. Market timing transactions reduce investment risk; and,
3. Market timing transactions increase investment returns.
At the limit, market-timing strategies eschew the diversification of a balanced portfolio approach in favor of
concentrating asset positions into a single capital market (stocks, bonds or cash). However, if the market
timing call is incorrect, the effect on portfolio value can be catastrophic. This is easy to see when the market
timing recommendation calls for movement from all cash to all stocks. Given the susceptibility of a 100%
equity position to unanticipated economic shocks, such a concentrated bet demands a high level of
confidence in forecasting skills. When a timing recommendation calls for abandoning stocks for cash,
however, the risk may be less easy to see. In this case, however, the catastrophe to wealth occurs not in the
form of loss of principal, but in the form of opportunity coststhe cost of missing the wealth-generating
process of the stock market.
The first important inquiry into market timing abilities is the Treynor and Mazuy essay published in
1966.126 The authors test the hypothesis that market-timing skill can be found in the universe of professional
mutual fund managers. They define market-timing skill as the ability to raise the sensitivity of the portfolio to
the return of the stock market prior to the onset of bull market periods and lower portfolio sensitivity to stocks
in anticipation of bear markets. Statistically, they compare (regress) returns in excess of the risk-free rate for a
mutual funds portfolio with returns in excess of the risk-free rate achieved by the stock market. If there is
evidence of successful market timing ability, the characteristic line of the regression equation (i.e. Beta) should
evidence a steep slope as the excess returns of the stock market grow large and a shallow slope as the excess
returns turn negative (i.e. the market earns less than a T-Bill). Upon evaluating the professional management
of 57 mutual funds over the period 1953 through 1962, the authors identify only one fund that exhibits
statistically significant ability to time markets successfully.
In 1975, future Nobel Prize winner William Sharpe proposed another approach to measuring the market
timing ability of investment professionals.127 Sharpe assumes that a manager changes the composition of his
and Discount Factors, Review of Financial Studies (fall, 1988), pp. 195-228 and Fama, Eugene F. & French Kenneth R.,
Business conditions and expected returns on stocks and bonds, Journal of Financial Economics (November, 1989), pp.
23-49. Whether investment managers can successfully exploit market predictability to earn excess profits remains an
open question. See, for example, Malkiel, Burton G., Can Predictable Patterns in Market Returns be Exploited Using Real
Money? The Journal of Portfolio Management (30th Anniversary Issued, 2004), pp. 131-141.
126 Treynor, Jack & Mazuy, Kay, Can Mutual Funds Outguess the Market? Harvard Business Review (July-August, 1966),
pp. 131-136.
127 Sharpe, William F., Likely Gains from Market Timing, Financial Analysts Journal (March-April, 1975).

69

or her portfolio based on market forecasts. Shifts in portfolio composition and weighting are, therefore,
proxies for the managers market predictions. A close correspondence between the predictions of the
manager and the actual direction taken by the market is evidence of superior market timing skill. However,
given the fact that markets tend to outperform risk-free investments approximately two-thirds of the time, a
market timer who is an eternal optimist will exhibit a 67% success rate. Sharpe therefore proposes several
statistical adjustments to measure the proportion of correct timing calls in both bull and bear markets. A
perfect market timer generates a score of 200% (correct prediction of each bull and each bear market) while
an eternal optimist with no prediction skills generates a score of 100% because he or she will always fail to
predict a bear market but will never miss a bull market. Sharpes research leads to two important conclusions:
1. There is little evidence of superior market timing skills among the population of professional
investment managers (i.e. the scores do not statistically differ from 100%); and,
2. The onus of transaction costs and commissions incurred in a simple one-time-per-year market
timing system between stocks and T-bills demands that the market timer make correct calls at a
74% frequency rate (i.e. achieve a score of 148 or better) to beat a naive buy and hold strategy.
Further refinements in the methodology of statistical measurements generated a sequence of market
timing studies in the 1970s and 1980s. The majority of these studies conclude that, in general, the
professional money management industry possesses negative market timing skills.128 The Journal of Financial
Services Research published, in 1998, a study that extends research on market timing abilities to an
evaluation of bank common (pooled) trusts during the period 1984 1992.129 The authors conclude that,
considered in the aggregate, bank trust department portfolio managers are unable to time the market
successfully by changing their portfolio betas in anticipation of differential market conditions and, thus, are
unable to outperform a passive buy and hold investment strategy. These results are what you would expect to
find in relatively efficient markets.130 These are markets in which the effects of economic, political, tax &
regulatory, and firm-specific news are quickly impounded in the price of stocks. In this type of market, the
price of any asset reflects the consensus opinion of investors regarding all relevant information affecting the
risks and rewards of owning the security.

PERFORMANCE OF ACTIVE MANAGERS

Research suggests that active managers find it difficult to earn abnormal profits (i.e., profits in excess of
expected returns, given the portfolios risk level). The performance of major pension and endowment funds,
and publicly traded mutual funds, should provide significant information. Major institutions attract above
average money management talent. Publicly traded mutual funds operate under the spotlight of daily
See, for example, the survey in Reilly, Frank K., & Brown, Keith C., Investment Analysis and Portfolio Management Fifth
Edition (Dryden Press, 1997), pp. 1015-1016; and Jones, Charles P., Investments: Analysis and Management Eighth
Edition (John Wiley & Sons, Inc., 2002), pp. 303-304.
129 Sahu, A., Kleiman, R., & Callaghan, J., The Timing and Stock Selection Abilities of Bank Funds: Evidence Based on
Meta-Analysis, Journal of Financial Services Research (1998), pp. 137-152.
130 The notable exception to the preponderance of academic opinion is found in Grinblatt, M. & Sheridan, T., Mutual Fund
Performance: An Analysis of Quarterly Portfolio Holdings, Journal of Business (1998), pp. 393-416. This study finds
evidence of market timing performance persistence and abnormal returns from market-timing strategies. However, the
magnitude of abnormal returns was not great enough to justify the costs of implementing timing strategies. Other studies
(e.g. Wagner, Jerry C., Why Market Timing Works, Journal of Investing (Summer, 1997), pp. 78-81) provide evidence of
positive returns to market timing for only limited sample periods. The phenomenon of limited periods of success for market
timers, however, has been more deeply examined (e.g. Bauer, R. & Dahlquist, J., Market Timing and Roulette Wheels,
Financial Analysts Journal (January/February, 2001), pp. 28-40) and the study reaffirms academic conclusions regarding
the low probability of market timing success.
128

70

published investment results. Successful funds attract millions of dollars in new contributions, while lagging
results can shrink funds rapidly as investors bail out.

Mutual Fund Performance

An early study of actively managed mutual fund performance that surveyed returns from 115 funds from
1955 through 1964 found no evidence of a consistent ability to achieve superior performance. Indeed,
performance was found to be worse than predicted by a 50/50 chance model.131
A comprehensive study of actively managed mutual fund performance from 1965 to 1984 (143 funds) by
Elton, Gruber, Das and Hlavka132 found that the mean annual alpha of funds evaluated for this period was
negative (-1.59).133 That is, the value added by active management, as compared with the performance of the
relevant benchmark index, was -1.59% per year. Although there were 34 funds with positive alphas during the
period, no alpha value was statistically significant. However, of the 109 funds with negative alphas, 21
negative results were statistically significant.
Even more dramatic was the finding that significant levels of active management (as evidenced by levels
of trading - i.e., portfolio turnover percentage) detracted from investment performance rather than added to it.
The following graph illustrates the studys findings. The x-axis measures the excess return (alpha) achieved by
active management.

Figure 4.1 Effect of Trading Activit y on Portfolio Return

1965 through 1984
Turnover
72% t o 162%

-2. 21%

High

51% t o 72%

-1.87%

34% t o 51%

-2.17%

22% t o 34%

-1. 11%

-2.50%

-2.00%

- 1.50%

-1.00%

- 0.50%

0.00%

Alph a
S o u rc e : E lt o n, E dwi n J ., Gruber, M art in J ., D as , S anjiv , and Hlav k a, M at t hew ,
E ff ic i enc y w it h C os t ly Inf o rm at io n: A R eint erpret at ion of E v idenc e fro m M an aged P o rtf o lio s , T he J o urnal f o r Fi nanc ial S t udies , 1993

Jensen, Michael C., Risk, the Pricing of Capital Assets, and the Evaluation of Investment Portfolios, Journal of
Business 42, no. 2 (April, 1969) pp. 167-247.
132 Elton, Edwin J., Gruber, Martin J., Das, Sanjiv, and Hlavka, Matthew, Efficiency with Costly Information: A
Reinterpretation of Evidence from Managed Portfolios, The Journal for Financial Studies, 1993
133 Alpha is defined as excess risk-adjusted return.
131

71

Low

Under 22%

-0.58%

Although active management subtracted value in every portfolio turnover range, the lower turnover
managers outperformed those with high turnover. A similar relationship was discovered between fund
expenses and performance results. Funds with high expense ratios do not generate enough extra return to
overcome the burden of the added expense. Good performance is negatively correlated with high-priced
management. This evidence strongly suggests that one key to long-term investment success is to keep
expenses low and to eschew trading oriented systems.

Money Manager Performance

If the managers of publicly traded mutual funds find it difficult to beat the market, what is the record of the
private money management industry? The Brookings Institution in 1992 published a comprehensive study of
private money management.134 This study utilized the proprietary SEI database (a private company
specializing in evaluating manager performance), which contains a wealth of information on private money
managers performance, total funds under management, accounts gained and lost over specified time periods,
fee schedules, equity share turnover, investment style, and so forth.
After adjusting for risk levels, the results for each rolling three year evaluation period 1983 through 1989
were as follows:

Figure 4.2 Equity Managers vs. S&P 500

S&P 500
Return

Interval

Active Management
Return

Percent
Underperforming

1983 - 1985

19.8

17.4

65%

1984 - 1986

18.5

17.4

57%

1985 - 1987

18.1

17.7

51%

1986 - 1988

13.3

13.0

54%

1987 - 1989

17.4

16.4

60%

According to the Brookings Institution, not only do the majority of private managers fail to beat an
unmanaged index, but there is no consistency of performance which would indicate long-term superior
management ability. Previous years best performers segregated themselves almost exactly according to
random chance during the following year. Indeed, by selecting the worst performing managers of the base
evaluation year, one would have had a slightly better chance of benefiting from top quartile performance in the
following year:

Figure 4.3 Prior Year Performance as an Indicator of Subsequent

Performance
Subsequent Year Performance

Prior Year
Performance

Top Quartile

Second Quartile

Third Quartile

Bottom Quartile

Top Quartile

26%

24%

23%

27%

Second Quartile

20%

26%

29%

25%

Lakonishow, Josef, Shleifer, Andrei, and Vishny, Robert W., The Structure and Performance of the Money Management
Industry. Brookings Papers on Economic Activity: Microeconomics, Brookings Institution, Washington, D.C. (1992), pp.
339-391.
134

72

Third Quartile

22%

28%

26%

24%

Bottom Quartile

32%

22%

22%

24%

The study concludes that active manager trade strategies are unproductive when compared to nave buyand-hold portfolios:
...trades made by the funds were counterproductive, costing on average forty-two basis points
relative to a portfolio frozen for six months and seventy-eight basis points relative to a
portfolio frozen for twelve months.135
There was no positive relationship between the fees charged by private managers and actual performance
results. When management fees are considered, the results from the search database would lead one to
conclude that active management subtracts value.136
SURVIVORSHIP BIAS
Burton Malkiels comprehensive study of mutual funds from 1971 to 1991137 points out that most studies
of mutual fund returns overstate the benefits of active management because of survivorship bias.
Suvivorship bias occurs when reseach evaluates only the track record of funds that have survived for the entire
period. Presumably, the failed funds went out of business because of poor investment performance. By failing
to include the record of failed funds, the resulting bias in performance measurement is substantial:

Ibid., p. 354.
Ibid., p. 351.
137 Malkiel, Burton G., Returns from Investing in Equity Mutual Funds 1971 to 1991, The Journal of Finance (June, 1995),
pp. 549-571.
135
136

73

Figure 4.4: Estimates of Mu tal Fun d Survorsh ip Bias

As Measu red by In vestmen t Returns
1971 through 1991
17.52%

1 7.09%

15.69%

All Equity Mutal Funds

Surviving Fu nds

S&P 50 0 In dex

S o u rc e: M alkiel , B urt on G ., Ret urns f rom Invest ing in Equit y M ut ual F unds 1971 to 1991, T he Journal o f Fi nance (June, 1995), pp. 549-571.

PERFORMANCE CONSISTENCY CASE STUDY: THE FORBES MAGAZINE HONOR ROLL

Persistence occurs when an investment manager produces returns consistently above or below the
average performance for a group of similar funds or for a comparable manager peer group. More formally,
performance persistence is a postive relation between performance ranking in an initial period and a
subsequent period. Investors are interested in the issue of performance persistence because, if there is a
strong performance correlation over time, investors can use past performance as a guide to predicting future
investment performance.
Malkiels study is an important study of the issue of performance consistency. Ideally, the investor would
like to identify funds with track records evidencing both high returns and consistency. Historical success is
meaningful to a prospective investor only to the extent that he can reasonably expect that it will continue.
Malkiel studies the Honor Roll of mutual funds published yearly by Forbes Magazine:
To earn a place on the honor roll, a fund not only had to have an extraordinary long-run
performance recordbut also had to meet certain consistency goals. Performance is
measured in both up and down markets, and funds must be at least top-half performers in
down markets to qualify for honor status. Thus, the Forbes method guards against the
selection of only high Beta funds following a sharp rise in the overall market. It is interesting
to ask if investors could have achieved superior returns buying these consistent
performers.138
The following graph illustrates Malkiels findings:

138

Ibid., p.566.

74

Fi gu re 4.5: Forbes' Hon or Rol l vs. S& P 500

5-Year Total Return of Honor Roll Funds
5-Year Total Return of S&P500 Index
151.73 %

11 3 .6 4%

1 12 .72 %
103.24%

1 0 3.6 6%
85 .28%

86 .91 %

7 2.1 5%
5 8.4 5%

3 4.4 9 %

1983-1987

1984-1988

1985-1989

1986-1990

1987-1991

S o u rc e: M alkiel , B urt on G ., Ret urns f rom Invest ing in Equit y M ut ual F unds 1971 to 1991, T he Journal o f Fi nance (June, 1995), pp. 549-571.

Malkiel concludes:
Most investors would be considerably better off by purchasing a low expense index fund, than
by trying to select an active fund manager who appears to possess a hot hand. Since active
management generally fails to provide excess returns and tends to generate greater tax
burdens for investors, the advantage of passive management holds.139
Academic studies have evaluated active management performance over approximately forty years, and
most have concluded that it is extraordinarily difficult for active management to add value consistently.140 A
comprehensive report by the Funds Management Research Centre reviews over 100 research papers
published globally on the issue of the persistence of performance in managed funds.141 The report concludes:
1. Good past performance seems to be, at best, a weak and unreliable preditor of future good
performance over the medium to long term. About half the studies found no correlation at all
between good past and good future performance. Where persistence was found, this was more
frequently in the shorter-term, (one to two years) than in the longer term.
Ibid., p. 571.
One performance study concludes that, on average, active mutual fund managers are, in fact able to select stock
portfolios that consistently outperform relevant comparative benchmark portfolios. However, once returns are adjusted for
cash (nonstock) holdings, expenses and transactions costs, their net returns underperform the market by one percent.
Wermers, Russ, Mutual Fund Performance: An Empirical Decomposition into Stock-Picking Talent, Style, Transactions
Costs, and Expenses, The Journal of Finance (August 2000), pp. 1655-1695.
141 Allen, David; Brailsford, Tim; Bird, Ron & Faff, Robert, A Review of the Research on the Past Performance of Managed
Funds, Funds Management Research Centre (June, 2003).
139
140

75

2. More studies seem to find that bad past performance increased the probability of future bad
performance.
3. Where persistence was found, the out-performance margin tended to be small. Where studies
found persistence, some specifically reported that frequent swapping to best performing funds
would not be an effective stgrategy, due to the cost of swapping.
Plausible explanations for these conclusions, in the authors opinion, include:

Methods that work well in one set of market conditions will not work well in new future
economies;

Fund managers, seeking to emulate the performance of their successful competitors, will copy
investment methods and/or poach investment staffs;

Large inflows of money to successful funds makes it difficult to find profitable new investments
and to maintain relative performance;

Future investment returns are difficult to forecast accurately and a significant portion of a funds
past performance may be attributable to random luck.

Interested readers may find a review of many articles on the performance evaluation of actively managed
funds, investment advice newsletters, television pundits, etc. in previous issues of Investment Quarterly and
Fiduciary Forum (www.schultzcollins.com) available on our firms website.142 The website also includes an
indepth paper on the topic of investment manager selection and retention policy [Without More: Trust
Investment Manager Selection and Retention Policy]. This paper provides an update on academic research
with special emphasis on studies suggesting that active managers can provide the opportunity to earn excess
profits. 143
EVALUATING ACTIVE MANAGER PERFORMANCE
Although most unbiased studies indicate that capital markets are efficient, and that it is difficult to beat
the market without assuming a correspondingly larger amount of investment risk, statistical analyses
demonstrates that a small percentage of active managers consistently add value, after expenses. However,
from the universe of thousands of active managers, random chance alone will produce some who fall into this
elite category. Therefore, decisions to incorporate active management in an investment portfolio require
statistical verification that the track record of the fund under consideration is both economically and
statistically significant, i.e., is attributable to manager skill rather than luck. Further, actively managed funds
should be regularly reviewed to determine their continued suitability for the portfolio.
Actively managed funds face a difficult burden of proof. In order to achieve returns in excess of a
comparable benchmark, they must:
Be able to forecast consistently and correctly those securities that offer better than average
returns;
Pay for their research costs from the returns that are actually generated;
Implement their buy and sell decisions in a cost effective manner; and
For example, Fiduciary Forum Vol. 5, #2 (September, 2001) offers an in-depth discussion of several articles: Chevalier,
Judith & Ellison, Glenn, Are Some Mutual Fund Managers Better Then Others? Cross-Sectional Patterns in Behavior and
Performance, The Journal of Finance (June, 1999); Jain, P.C. & Wu, J.S., Truth in Mutual Fund Advertising: Evidence on
Future Performance and Fund Flows, The Journal of Finance (April, 2000) and Zheng, L., Is Money Smart? A Study of
Mutual Fund Investors Fund Selection Ability, The Journal of Finance (June, 1999).
143 http://www.schultzcollins.com/node/442
142

76

Avoid concentrating their bets to the extent that investor risk is significantly magnified.
Fortunately, there are a number of straightforward statistical tests that measure a managers forecasting
ability. These tests represent a set of diagnostics to determine whether proprietary investment strategies are
likely to add or subtract value. Employing investment strategies leading to extreme levels of asset
concentration, without performing appropriate diagnostics within the money management organization,
however, may be evidence of imprudent asset management. Such conduct puts the organizations interest in
collecting fees above the clients interest in achieving successful financial outcomes. In most respects, acting
in the capacity of investment advisor or money manager without prudent diagnostics and internal controls is
no different than selling medications without sufficient research and testing and without sufficient quality
control in the manufacturing and distribution of the pharmaceutical product.144
An especially important set of statistical diagnostics is the measurement of forecast errors. If a portfolio
manager has perfect forecasting ability (a prophet), diversification would be a stupid and wasteful use of client
money. He or she would simply own the single security that over the forthcoming planning horizon generates
the greatest return. If a portfolio manager has forecasting ability that is less than perfect, the optimal number
of securities that should be held within the portfolio exists on a spectrum that extends from only a few
securities to a large number of stocks. As the forecasting ability approaches 50/50, the portfolios
composition should approach the fully diversified index or customized asset benchmark that aligns with the
liabilities to be discharged by the portfolio. Managers who market time by overweighting or underweighting
sectors must have exceptionally high levels of forecasting skill because their portfolios tend only to own a few
stocks concentrated in a few industries.
Additionally, it is worth spending a few moments to consider the marketing of a disciplined investment
philosophy by many money management firms. Goldman Sachs Abby Joseph Cohen notes: discipline
sometimes does not give the right answer. It just gives a formulaic answer and can intensify the
consequences of an incorrect answer.145 There is a critical difference between being a disciplined investor
(i.e. staying the course, not making common mistakes, etc.) and being a professional investment firm capable
of adding value to a benchmark. Although the two propositions sound similar, the first claim merely suggests
that the organization will not blunder as badly as amateur investors; the second that the organization
possesses unique advantages that enable it to outperform its professional competitors.146

Martin Leibowitz, a managing director at Morgan Stanley, pointing out the folly of relying on past track record as a guide
to future results, stresses the need for internal diagnostics as a necessary condition for prudent investment management.
Leibowitz suggests rephrasing the prospectus warning on past performance: A more ominous rephrasing would be, Past
performance is not even a good guide to the quality of the decisions that went into that past performance. Yet, the
ultimate issue is the soundness of the decision process itself: Was all knowable information incorporated? Was the
reasoning thorough and sound? Were alternative scenarios considered and contrary views sought? Was a well-planned
implementation and monitoring program establishedand then followed? Was there a routine postmortem analysis of
lessons learned? Leibowitz, Martin L., Alpha Hunters and Beta Grazers, Financial Analysts Journal (September/October,
2005), p. 34.
145 Cohen, Abby Joseph, Aristotle on Investment Decision Making, Financial Analysts Journal (July/August, 2005), p. 29.
146 Larry Harris, chief economist for the U.S. Securities and Exchange Commission, explains the problem as follows:
Traders who estimate values from the same information, using the same methods, tend to estimate the same values.
Their estimates are highly correlated. They must compete with each other to profit from their insights. Traders whose
estimates are not closely correlated with the estimates of other traders have orthogonal estimates (Orthogonal comes from
a Greek word that means at right angles.) Traders obtain orthogonal estimates of value when they base their estimates
on information that other traders do not use or when they analyze data using different methods than other traders use.
The most profitable traders have very accurate estimates of value that are uncorrelated with the value estimates made by
other traders. Harris, Larry, Trading and Exchanges: Market Microstructure for Practitioners, (Oxford Univ. Press, 2003), p.
237. Thus, a professional money manager can expect to beat the market only if he or she possesses high forecasting
accuracy and the managers forecasts deviate from the consensus forecasts of other market participants. It is hard to beat
the market; and an organization should not claim that it is likely to do so prior to confirming its abilities and prior to
charging fees to the public.
144

77

Thus, the essential question is: what makes the money manager believe that its disciplined approach can
create excess profits (profits beyond those reasonable for the risk to which it exposes client wealth)? The
question is critical because without a verifiable answer, the investor should have no expectation that trades
will be profitable. Organizations that do not, in fact, possess true competitive advantages that allow them to
generate excess profits (i.e., add value for their clients) tend to emphasize qualities like discipline, personal
service, enthusiasm for meeting organizational goals, and so forth in their sales and client communications
materials. It is not enough, in a highly efficient and competitive market, merely to advance reasons why an
investment strategy should work. One must also be clear on why and how other sharp-pencil institutional
trading organizations will lose when faced with your organizations resources and skills.147 Professional money
managers who are merely better than average (i.e. better than the average individual investor), will earn less
than average returns in the market.
One danger of focused portfolios lies in the fact that active management offers only a conditional
expectation of success. Return is conditioned on the portfolio managers forecasting abilities and trading
skills. Indexed or asset-class investing, however, offers unconditional return expectations because the
investor has a positive and unconditional expectation that he or she will earn the risk-premium of the capital
market. The investor does not always attain the expected premium but lacking a positive expectation for
reward, only risk-free investments would remain in the marketplace. Thus, a portfolio that avoids active or
conditional investment risk is one that is well-constructed, broadly diversified, and that aligns with the
investors risk/return requirements.148
It is important to understand that active investment management may, in fact, represent a prudent course
of action for investors. In the context of this discussion, two points are of interest because they reflect the
ongoing debate over the wisdom of selecting active investment managers:
1. The focused portfolio school of thought argues that implementing a benchmarked portfolio
(benchmarked to liabilities or to an asset-side allocation only) is the risky gamble. Indexes, in this
view are capitalization-weighted vehicles that force investors to buy large portions of highly priced
stocks and small portions of stocks that may represent potential bargains. Risk is avoided by
deviating from the benchmarks (i.e., making active manager decisions) so that you stand a better
chance of making money; and, on the other side of the argument,
2. Statistical tests demonstrating that the active manager adds positive risk-adjusted value for the
benefit of the investor may not be sufficient to justify placing wealth in the hands of the manager.
This is because, the investor could have achieved an unconditional return without active manager
risk and; therefore, to justify assuming the extra risks and costs the investor requires some
amount of positive alpha merely for taking benchmark risk. A second level of testing is required
to determine, given the investor risk aversion, whether the amount of positive value added by the
manager is sufficient to justify the extra risks.
A case can be made for active management provided that the investor selects the managers carefully.149

Harris explains the concept of comparative advantage as follows: On average, better plays win games. Good players
and even great players do not generally win when they play against even better players. A player has an absolute
advantage when he or she can do something well.A 2:20 marathoner will win the vast majority of marathons that are run
every year. Such a time, however, would have been good for only 36th place in the mens marathon at the 2000
Olympics. To win a game, you must not just play it well. You must play it better than your opponents. p. 476.
148 Siegel, Laurence B., Benchmarks and Investment Management (The Research Foundation of AIMR, 2003).
149 In some cases, it would be demonstrably imprudent not to select active management strategies. One example is the
decision to utilize cash matching or immunization strategies when managing towards a fixed income liability cash flow
stream. It would be highly unlikely that the characteristics of indexed investment products would match the characteristics
of the funding liabilities.
147

78

PASSIVE MANAGEMENT
Passive funds usually attempt to mirror the returns and risks of an index or an asset class. There are
several broad categories of passively managed investments.

Index Funds
An index is an artificial indicator of price levels in a market segment. To build an index, securities are
grouped together based on their exhibition of certain quantifiable characteristics, and the weighted average of
their prices is computed on a daily basis. The weighting is determined by, for example, their relative market
capitalization (the Wilshire 5000 index tracks the stock performance of the 5000 largest companies; the S&P
500 is an index of five hundred large firms from a cross-section of representative U.S. industries); or, by
composition (the U.S. Treasury Bond Index is composed exclusively of Government Securities; the Corporate
Bond Index is composed exclusively of corporate debt issues).
Index funds seek to replicate a stock or bond index. They may buy every security in the index, or a
representative sample of securities whose behavior mimics the index (sampling or sensitivity indexes). The
fund manager makes no forecasting decisions. Management attempts to replicate the market rather than to
beat it.

Full Replication Index Funds

These index funds hold most or all of the securities contained in the asset class benchmark, in the same
weightings that exist within the benchmark. Purchase and sale of individual securities are based on changes in
their relative market capitalization weights.

Sample Index Funds

These index funds hold representative samples of the securities contained in the benchmark. Sample
index funds built through a random sample process often exhibit large tracking error vis a vis the benchmark
index. Sample index funds built through a stratified cell approach minimize tracking error. In a stratified cell
approach, the risk/return characteristics of underlying securities are decomposed and quantified. Each cell
represents one such characteristic, and securities are selected on the basis of how closely their composition
reflects the required characteristic.

Other Index Investment Approaches

There are several other approaches used by index funds. In general, these approaches seek to provide
performance superior to the index, while retaining the objectivity and risk characteristics of the index
approach. These include optimization indexes, such as stratified cell index funds; enhanced index funds,
which generally use derivatives in an attempt to benefit from market mispricing; and index funds which apply
different market weightings to component securitiesequal weighted indexes, fundamental indexes, etc.

Structured Asset Class Funds

A Structured Asset Class fund is a group of securities that exhibit comparable risk/return characteristics.
These funds usually decompose an index to capture a specific dimension of risk or return (e.g., the value
style subset of S&P 500 stocks with low market value to asset value ratios); or, may group securities into a
unique index reflecting certain historical risk/return characteristics (e.g., funds based on yield curve
positioning).
Structured Asset Class funds capture returns by

79

purchasing all securities with comparable risk/return

characteristics along an identifiable investment dimension (e.g., market size, yield curve placement, etc.). They
may or may not try to track a benchmark index. Unlike an actively managed fund, the fund manager makes no
forecasting decisions. Management seeks neither to beat the market nor, strictly speaking, to replicate it. The
primary objective is to capture the long-term returns that flow from exposure to risk factors specific to an asset
class. Purchase and sale of individual securities are based on passive filters (see below) designed to preserve
the stated risk/return characteristics of the fund.

Structured Asset Class: Equity Fund Management

Whereas a large company stock index fund might own every security used to calculate its benchmark
index, a small company structured asset class fund might impose passive inclusion or exclusion filters on the
universe of securities. Inclusion filters often mandate purchase of securities with certain accounting ratios
within a specified range, or purchase of target securities at prices below the bid/ask spread. Exclusion filters
might proscribe ownership of bankrupt firms, or initial public offerings.

Structured Asset Class: Fixed Income Fund Management

Fixed income index funds might own a bond through all yield curve environments simply because the
weighting of the index demands it. A structured asset class fund investing in fixed income, on the other hand,
may shift maturities based on available yields (as reflected in the slope of the current yield curve) and on a
horizon analysis of total expected return over the relevant holding period. Such an approach eschews
forecasting (whether of interest rates, inter-sector spread or credit quality) because all analytical inputs are
derived from the current yield curve environment.

EVALUATING PASSIVE FUND PERFORMANCE

Index Funds
When evaluating index funds, the investor is primarily interested in how well the fund matches the risks
and rewards of its comparable benchmark index. For example, how closely does the Vanguard 500 Index
Trust match the returns of the S&P 500 US Stock Index? Is the investor getting what he or she paid for?
No index fund tracks its benchmark with absolute precision [funds have expenses; benchmarks are merely
paper portfolios]. Additionally, there is no generally accepted standard for determining whether an index fund
remains prudent or suitable for an investment portfolio. Rather, the function of an evaluation is to present a
range of credible and relevant information so that, taking the weight of the evidence as a whole, the investor
can formulate intelligent judgments as to past performance as well as to likely future performance. Such a
judgment is relative. Are there better investment vehicles? Is it worth incurring expenses to make changes?
Are positive or negative trends likely to continue? Does the index fund offer benefits in some areas sufficient
to outweigh negatives in other areas?150
Generally, two methods may provide important insight into index fund performance:

150

Descriptive Statistics. This method of quantitative analysis is interested in the shape of the
benchmark indexs distribution of returns. What is the average return, the degree to which returns
cluster around the average, and the degree to which the benchmark index generates extreme returns?
Once the investor understands the risk and return characteristics of the benchmark index, he can
determine how well the fund captures critical risk/return dimensions.

Collins, Patrick J., Monitoring Passively Managed Mutual Funds, The Journal of Investing (Winter, 1999), pp. 49-61.

80

Statistical Correspondence. Also known as Regression Analysis, this method plots the period-byperiod fund returns against those of the benchmark index. Ideally, the benchmark and fund returns
match perfectly. The extent to which the monthly return data points plot off the line indicates the
magnitude and prevalence of tracking error. A high degree of tracking error indicates that the index
fund is not successfully capturing the risks and returns of the indexi.e., investors are not getting
what they paid for.

Structured Asset Class Funds

Structured asset class funds are passively managed funds that incorporate asset management strategies
sometimes found in actively managed funds. The most prominent manufacturer of structured asset class
funds is Dimensional Fund Advisors (DFA). Dimensional funds generally avoid security selection decisions
based on macro-economic forecasting, industry analysis, or security valuation models. Thus, unlike most
actively managed funds, Dimensional does not generate returns through price change forecasting or market
timing activities. Additionally, unlike many actively managed funds, Dimensional funds avoid asset
concentration in only a few stocks or bonds in favor of owning a broad selection of securities within the
applicable category (US large, US small, etc.). Hence, the term asset class funds they own many of the
eligible securities within the asset class.
Dimensional funds are structured by virtue of the application of certain passive filters or screens that are
used to eliminate securities with undesirable characteristics. Thus, unlike most funds, investment decisions
remove from rather than select for. Filters are passively applied to the universe of eligible securities. For
example, the index of US small stocks may be filtered to eliminate companies in bankruptcy, companies
without sufficient market liquidity, companies that are primarily closely-held, IPOs, and so forth. Generally, all
securities remaining after application of the filters are purchased on a capitalization-weighted basis. Thus, a
Dimensional equity fund may look very much like a full-replication index fund that has been swept to
eliminate securities with certain undesirable characteristics. The filter rules for fixed income (bond) portfolios
often take the form of decision rules based not on interest rate or yield-curve forecasts, but on identifying the
position on the current yield curve that offers the most favorable credit or maturity risk spreads.
Many of Dimensionals equity funds are actively managed with respect to their market execution
strategies. Unlike index funds that must present buy and sell orders quickly in order to avoid drifting away from
their underlying index, structured asset class funds are not overly concerned with avoiding tracking risk (risk
that the returns will differ from the index). As patient traders they may present transactions more slowly and,
therefore, be less subject to unfavorable bid/ask spread expenses or market impact costs. Additionally,
executing off the market trades through use of electronic communications networks that allow large
institutions to deal directly with each other rather than through an exchange or broker/dealer intermediary,
can reduce trading commissions and spreads. Dimensional tries to add value through application of filter
rules and trading strategies rather than through price forecasting and market timing.
When evaluating structured asset class funds, the investor is primarily interested in whether the filter rules
and trading strategies have added value vis--vis the relevant benchmark index. In some cases, adding value
also extends to the concept of preserving value after fees and expenses. Many Dimensional funds, for
example, operate in environments known for high liquidity costs. In these environments, underperforming a
zero-cost paper index by only a few basis points per month is a major achievement.

81

C H A P T E R S I X : P O R T FO L I O M A N AG E M E N T

A fully integrated portfolio management must consider the asset allocation issue from the unique
perspective of the individual investor. For example, changes in personal wealth often alter the investors risk
tolerance.151 Some investors, if a portfolio increases in value, will be better able to tolerate risk (the increase
in value acts as a cushion against downturns). Others will become more conservative because portfolio gains
make it possible to bear less risk in order to achieve financial objectives. Changes in risk tolerance are highly
personal because each investor has a unique sensitivity to changes in wealth. Thus, a fully integrated portfolio
management system:
1. Optimizes investor utility by creating an initial asset mix that generates appropriate risks and
rewards ; and,
2. Monitors changes in investor circumstances to assure that the portfolio continues to be aligned
with the investors purposes, distribution requirements, and financial circumstances.
ASSET ALLOCATION APPROACHES
As a practical matter, asset allocation can follow one of several strategies:

Tactical Asset Allocation

Tactical asset allocation assumes that the investors risk aversion is fixed. Consequently, wealth changes
do not affect allocation decisions. However, asset price changes do affect capital market expectations (future
expected returns, risks, and market correlations). Several well-known portfolio management approaches flow
from this viewpoint. At the extremes, it leads either to strict contrarian strategies, or to market timing systems.
Strategic Asset Allocation
Strategic asset allocation determines the investors desired exposure to sources of systematic risk. It is
based on the belief that capital markets are generally efficient over long-term planning horizons. An efficient
market rapidly incorporates all known information regarding economic and financial matters into the current
price of a security. Thus any current asset price represents the consensus opinion of the assets value. There
is rarely hidden value waiting to be discovered by an astute investor. Price changes occur because:

New information causes a reassessment of the assets value; or

There are more liquidity sellers than information buyers (but this results from random
chance; no investor has the inside track on the direction of price change.

Under this theory, then, the prices of securities are exactly what the market requires to clear. Once the
asset allocation mix is set, the portfolio remains fixed for relatively long periods of time. Among the asset
management approaches that flow from this viewpoint are Buy-and-Hold and Constant Mix (which periodically
rebalances back to the initial asset allocation) strategies.

Insured Asset Allocation

Insured Asset Allocation integrates both tactical and strategic methods. It takes the efficient market
viewpoint that changes in price lack predictability sufficient to form profitable market-beating systems.
However, it assumes that changes in wealth affect investor risk aversion. As portfolio value declines, the
Riley, William B. & Chow. K. Victor, Asset Allocation and Individual Risk Aversion, Financial Analysts Journal
(November/December, 1992), pp. 32-37.
151

82

investor is assumed to become more sensitive to investment risks; as the value increases, the investor
becomes more comfortable with risk.
The Insured Asset Allocation strategy sets a floor on the dollar value of a portfolio. As the portfolio value
approaches the floor, assets are shifted to risk-free Treasuries. If the value sinks to the floor value, the entire
portfolio will consist of Treasury Bills and will suffer no further declines. Conversely, as portfolio value climbs
above the floor, more funds are committed to equities. Many Insured Asset Allocations have equity
commitments of two to four times the spread between current portfolio value and the floor value. This
difference is called a multiplier. 152

THEORETICAL PAYOFFS TO DIFFERENT ASSET MANAGEMENT APPROACHES

The following graph depicts the value of a portfolio (on the Y-axis) under three asset management
approaches as the value of the risky assets (on the X-axis) changes. We consider the Buy and Hold, the
Constant Mix and the Insured Asset Allocation approaches. The investment is made when the value of the risky
asset is $100. The Buy and Hold portfolio assumes an initial commitment of 60% risky assets to 40% T-Bills.
The Constant Mix assumes constantly rebalancing to 60% equity/40% T-Bill. The Insured Portfolio approach
assumes a floor value of 70 with a multiplier of two. Therefore, the initial equity investment position of the
Insured Portfolio is (100-70) x 2 = 60% equity/40% T-Bill. Although each portfolio starts with the same ratio of
equity to risk-free asset, each diverges in value as the price of the risky asset portion changes.

Figure 6.1 Comparative Performance of Asset Manag ement Styles

160.00
140.00
120.00
100.00
80.00
60.00
40.00
20.00
0.00
0

10

20

30

40

50

60

70

80

90

100 110 120 130 140 150 160 170

Value of Assets: Constant Mix

Value of Assets: Insured Portf olio

V alue of Assets: Buy & Hold

The white paper entitled Risk, Return & Rebalancing, located on the website at www.schultzcollins.com, contains a
detailed discussion of portfolio allocation strategies.
152

83

Buy and Hold

In the case illustrated above, the Buy and Hold investor placed 40% of the portfolio in short-term T-bills
and 60% in the risky market portfolio (a combination of stocks, bonds and real estate). What are the
implications?

The dollar value of the portfolio will not fall below that of the 40% commitment to TBills;

The portfolio has unlimited upside potential;

The future value of the portfolio is (approximately) linearly related to the performance
of the risky asset portion, with the rate (slope of the line) of future value change
equal to the proportion of the portfolio committed to the risky assets (in this case,
0.60).

Constant Mix
Investors whose risk aversion is not affected by wealth changes will employ a buy low/sell high strategy. If
an asset price declines, they will buy into the falling market. Conversely, they will take profits by selling into
rising markets. A typical example of this strategy is the Constant Mix management style. The Constant Mix
strategy restores the asset allocation to its original proportion of risk-free and risky assets as market prices
change. What are the implications?

Buying into declining markets while readjusting the equity portion to a constant
percentage of portfolio value means that, theoretically, 100% of the portfolio is
exposed to risk;

As markets increase in value, equities are sold. Therefore, the investment payoff will
tend to lag behind the payoff for a buy-and-hold portfolio that does not trim back
equities;

As equity markets decrease in value, low risk assets are sold and equities are
purchased to maintain targeted allocation levels. Therefore, in declining markets
returns will also tend to lag behind a buy-and-hold portfolio;

Therefore, the future change in value of the portfolio has a concave (turned down
curve) slope, which generally lies below the straight-line buy-and-hold payoff line.

Insured Asset Allocation

Investors whose risk aversion exhibits above average sensitivity to wealth changes may employ a buy
high/sell low strategy. This is a momentum driven strategy best characterized by the Insured Asset Allocation
portfolio style. What are the implications?

The portfolio is only exposed to risk on the amount above the floor value;

The rate of future change in portfolio value depends on the percentage of the
commitment to risky assets when the dollar value is above the floor value. On a
million dollar portfolio, for example, with a floor value of $800,000 and an risky asset
multiplier of 3, the equity commitment equals (1,000,000 - 800,000) x 3 =
$600,000;

84

As markets rise, the commitment to risky assets means that the investment payoff
will increase exposure to equities, creating a positive feedback loop with beneficial
effects on overall return;

As markets fall, the commitment to risky assets will scale back (by a factor of three in
the above example), creating a beneficial negative feedback loop until the value of
the portfolio reaches its insured value or floor;

Therefore, the investment payoff will be convex (a turned up curve) and will
outperform both the buy-and-hold straight-line payoff and the Constant Mix allocation
payoff.

Cursory analysis would seem to indicate that the insured portfolio asset allocation strategy is superior to
other approaches. Indeed, the strategys logic was so compelling that by the mid-1980s, four times more
professionally managed money used the insured portfolio system (buy high/sell low) than used the Constant
Mix (buy low/sell high) system. By October 1987, approximately four times as much institutional wealth was
managed under the Insured Portfolio approach than under the Constant Mix approach. When market prices
started to drop, 20% of the market wanted to buy while 80% of the market wanted to sell. Many economists
believe that this order imbalance was a primary contributor to the 1987 market crash. During the crash,
severe price declines eliminated the entire buy side of the market. Insured portfolios could not sell, and sank
far below their floor values. Thus, in a market crash, when portfolio insurance was needed most, it was
ineffective.
If either asset management strategy dominates the money management community, it sows the seeds of
its own destruction. Insured portfolio management creates market volatility, because its momentum driven
trading strategy exacerbates market swings. However, in volatile markets, it cannot provide the investment
payoff functions that it promises. Conversely, Constant Mix portfolio management reduces market volatility by
buying when prices are falling and selling as prices rise. But Constant Mix strategies require volatility (sudden
price reversals) to make contrarian bets worthwhile. Hence, one may reasonably conclude that there is no
best asset management approach, because each approachs payoff differs under varying market conditions.
The shape of the payoff curve for each strategy depends on equity market volatility. In trending markets,
the Insured Portfolio strategy buys into the favorable trend and rides the winners up. The more the winners
win, the better the portfolios performance. Conversely, the Constant Mix strategy pares back the winners (take
profits on the way up) to maintain the proportionate value of the fixed income anchor position. Therefore, when
markets trend up or down over long periods, this portfolio management approach has relatively poor
performance when compared to Insured or Buy and Hold styles. When market trends reverse, or are
nonexistent, however, results are reversed. The Constant Mix portfolio outperforms other approaches by taking
advantage of price changes.

IMPACT OF TRADING ACTIVITY ON PORTFOLIO RETURNS

Not only does each asset management approach have a unique set of payoffs; it also has its own pattern
of trading activity:

The trading activity required to maintain the Buy and Holdapproach is zero;

Periodic rebalancing requires only sporadic trading activity for the Constant Mix style,
and results typically in buying low and selling high;

Trading activity for the Insured Portfolio approach is a function of the momentum of
portfolio value change. The higher the price increase of the risky assets, the greater
the trading activity required to maintain the multiplier. Momentum-driven trading
tends to buy high and sell low.

85

Paper Portfolios and Real World Results

Portfolio management approaches must be implemented in the real world. Implementation entails trading
costs, liquidity costs and (except for qualified plan investors) taxes. One way to measure transaction costs is to
evaluate return differences between hypothetical and real portfolios. David J. Leinweber measured
implementation shortfall by tracking return differences between the paper portfolio recommended by the Value
Line rating service and the actual Value Line mutual fund that replicates the paper index. From 1979 to 1991
the Value Line paper index portfolio had a 26.2% annualized rate of return. The actual Value Line fund,
however, earned a net after expense return of only 16.1% during the period.153 The return difference measures
the (pre-tax) implementation costs.
At first, it seems incredible that implementation costs caused a live portfolios annualized returns to lag its
index by 10.1% over a thirteen-year period. Most people assume that implementation costs refer to:
1. Commissions paid to buy and sell securities in the marketplace; and,
2. Operating expenses associated with marketing and managing the live portfolio.
However, these costs are often insignificant relative to other hidden or implicit asset management costs.

Cost of Liquidity
As Wayne H. Wagner points out: many costs will be incurred long before the marketplace ever sees the
order.154 Wagners study measures market impact costs and how such costs affect a live portfolios value.
Market impact costs include a broad range of charges assessed against the portfolio by the financial markets
in exchange for providing trading liquidity. When a portfolio manager sells a stock, the sale is rarely to a
counterparty that wants to buy the precise number of shares offered. Instead, the portfolio manager must
trade with a liquidity provider.
For quote-driven markets such as NASDAQ, this is the role of the dealer; on the New York Stock Exchange,
this is the role of the exchange floor specialist. Specialists are required to trade the stocks in which they
specialize, at a price they determine, with any floor trader at any time. The specialist buys incoming stock
orders at a bid price; takes them into inventory; and resells them at a higher ask price. The specialist sets the
bid-ask spread to provide adequate compensation for the risk of holding inventory. Holding inventory ties up
capital if the stock cannot be resold quickly. The specialist must also cover the risk of absolute loss on
inventory if prices should plummet.
Dealer bid-ask spreads are dynamic. As the flow of buy or sell orders strains a dealers inventory, the
spread adjusts quickly, either up or down. The magnitude of the shift depends on whether the dealer is buying
increased inventory or selling surplus. The dealer must negotiate with stock traders whose spreads can be
many times greater than the dealers. Thus, when the dealers inventory is under pressure, he must transact
with traders who are under no obligation to buy or sell, and have no market making duties.
This is why even small market orders can have market impact. As a dealers inventory grows or shrinks, he
becomes more and more sensitive to pricing risks inherent in his net position. Therefore even small market
Leinweber, David J., Using Information from Trading in Trading and Portfolio Management, Execution Techniques, True
Trading Costs, and the Microstructure of Markets ed. K. F. Sherrerd (AIMR, 1993), pp. 25-26.
154 Wagner, Wayne H., Defining and Measuring Trading Costs, Execution Techniques, True Trading Costs, and the
Microstructure of Markets ed. K. F. Sherrerd (AIMR, 1993), p. 15.
153

86

orders can have a large marginal effect on the magnitude of the bid/ask spread. The next time you consider
purchasing a particularly hot security, remember that trading costs for momentum stocks can be astronomical:
The price obtained by the dealers customer depends to a large extent on how the customer
is trading relative to the crowd. Is the customer trading against the crowd, with the crowd, or
independently of the crowd?Think of transaction costs as an iceberg, with the commission
being the tip above the surface. The major parts of transaction costs are unobservable.155
As Wagner remarks: Market liquidity is not a free good. Those who absorb market liquidity must pay those
who supply it.156 Those who trade against the crowd are suppliers. They are reducing the specialists risk. The
crowd is increasing it. Specialists will pay liquidity suppliers for inventory risk reduction, and pass along the
cost (with a markup, of course) to the crowd. Liquidity suppliers are therefore able to sell at a relative premium,
and buy at a relative discount.

LIQUIDITY COSTS OF PORTFOLIO MANAGEMENT STRATEGIES

Liquidity costs have a profound effect on the different asset management approaches. The Insured
Portfolio approach generates significant trading activity. As risky asset prices fall, they are sold (in favor of TBills); as prices rise, the investor buys more stocks. Portfolio trading matches market momentum. Buy orders
are submitted at a time when buy orders dominate trading activity; and sell orders are entered when most of
the market wants to sell. Because trade orders are entered when liquidity is scarce, the Insured Portfolio style
must pay a premium price for trade executions.
What does it cost to demand liquidity when it is in high demand? Wagner and Edwards tracked 54,000
trades and concluded that brokerage commissions (the cost just to enter the order) paid per trade were 5.6
cents per share. When the trades reached the market, dealer/specialist bid-ask spread costs and market
impact costs deducted an additional 12 cents per share. Finally, the cost of immediate execution (i.e., the cost
of liquidity) deducted an additional 99 cents per share.157 Commissions thus represented only a small fraction
of total trading costs. By definition, the Insured Portfolio style entails momentum trading and, to protect the
downside floor, demands immediate execution. Clearly, the impact of trading costs on this strategy is
enormous.
Constant Mix portfolio management approach rebalances the portfolio periodically. Trading volume is
significantly lower than the Insured Portfolio style. More importantly, however, Constant Mix employs what
amounts to a contrarian discipline, selling assets as their prices rise and buying as prices fall. This means that
those who employ the Constant Mix strategy are liquidity providers, and are in a position to reap profits
therefrom. How much of a difference does this make?
Wagner and Edwards estimate that brokerage commissions paid by liquidity providers are the same 5.6
cents per share. However, supplying liquidity adds 4 cents of value when spread and impact costs are
measured, and 96 cents of value when liquidity costs are measured.
The implications of large trading costs are apparent:
1) Buying highly recommended or popular stocks (momentum or trend buying) is extremely expensive;
2) Trading stocks frequently is also costly. As Wagner states: As a whole, active management
performance falls short of index fund performance by between 100 and 150 basis points. Where does
Treynor, Jack & Wagner, Wayne, Implementation of Portfolio Building: Execution, Managing Investment Portfolios: A
Dynamic Process ed. J. Maginn & D. Tuttle, Warren, Gorham & Lamont (New York, 1990), pp. 12-1 to 12-50.
156 Ibid., p. 15.
157 Wagner, W. H. & Edwards, M., Best Execution, Financial Analysts Journal (January/February, 1993), pp. 65-71.
155

87

the money go? Into the frictional costs of getting security analysts and portfolio managers ideas into
the portfolio;
3) Recovering trading costs is difficult. As Charles Ellis estimates, the operating costs of the average
actively managed mutual fund amount to 1.6% per year. Over the long term, equity markets have
provided a 6 percent premium over the risk-free return. Thus an active fund manager must outperform
the 6% equity premium by 26.7% just to recover costs and break even with the market;158
4) Disciplined value style investing (i.e., buying out-of-favor securities) may afford opportunities for
trading strategies that are also portfolio profit centers.
5) The trading advantage goes to two groups of traders:
a) Information-based traders who act quickly and who possess information more valuable than the
heavy trading costs; and,
b) Passive investors with a value style orientation using periodic rebalancing.
The latter let the market to come to them. The passive investor pays no premium for speedy transactions,
since they are not needed. Empirical evidence suggests that passive, value-oriented portfolios best realize
trading cost advantages. These investors tend to purchase out of favor securities (i.e., trade against the crowd
on a buy low/sell high basis).
Theoretically, the Insured Portfolio asset management approach works best in a market that is strongly
trending. Conversely, the Constant Mix approach works best in a more volatile market characterized by
sudden price reversals.

TRADING DECISIONS, BEST EXECUTION AND LOSS OF INVESTOR WEALTH

Trading is anti-performance. This should mean that money managers have a strong incentive to control
trading costs. Paradoxically, however, this may not always be the case. There is evidence suggesting that
money managers are directing trades to venues that do not offer the most favorable trade execution services
(although the trades execute at the best Best Bid/Offer cost measurement criteria; and, theoretically fulfill the
requirements of best execution). For example, small trades for stocks listed on the New York Stock Exchange
may be sent to NASDAQ for execution. Selection of the NASDAQ venue may significantly reduce the probability
of a trader stepping in to offer price improvement [trading terms better than the Best Bid/Offer]. One possible
explanation for the propensity of the U.S. money management industry to direct trades to higher cost venues is
that the managers receive a bundled service package from brokers. In addition to pure trade execution,
money managers may also receive data access, research, computer systems, etc. in the form of soft dollar
compensation. One study indicates that over half of all U.S. institutional commissions are directed or
pledged in advance; and, that the recipients of the directed order flows compensate the money managers
through soft dollar arrangements.159
Regulatory agencies have expressed concern that such arrangements may be a breach of fiduciary
obligations to clients. The Department of Labor has been especially interested in this issue; and has made it
clear that commissions are a use of a retirement plans assets. Plan assets must be managed for the
exclusive benefit of plan participants and beneficiaries. Receipt of soft dollar compensation may represent
use of client/plan funds to pay for expenses associated with operating a money management firm. Use of
client funds allieviates the necessity for the money management firm to commit its own capital for business
expenses; and may represent a classic principal/agent conflict of interest.

Ellis, Charles D. Investment Policy, Irwin (Chicago, 1993), p.9.

Conrad, J., Johnson, K. & Wahal, S., Institutional Trading and Soft Dollars, Journal of Finance (vol. 56, 2001), pp. 397422.
158
159

88

Directed brokerage arrangements are often a hallmark of Wrap Fee accounts sold, primarily but not
exclusively, to retail investors. A Wrap Fee account is an all-in or bundled service package offered by a
brokerage firm to its customers. A single fee pays for all trade costs, custodial services, periodic reporting of
account positions, investment manager selection and monitoring, and performance reviews. Often the single
fee arrangement includes a written IPS, asset allocation advice, rebalancing, and other portfolio management
services. However, a brokerage companys wrap fee program may require exclusive trade execution through
the sponsoring broker. For money manager clients not in the wrap fee program, the manager is free to seek
any trade execution venue, including low-cost ECNs. In certain cases, money management firms may
sequence the wrap fee client trades last to avoid violating fiduciary responsibilities to other clients. That is to
say, the wrap fee clients may be purchasing securities at the tail-end of a buy order (the highest price) or
selling securities at the tail-end of a sell order (the lowest price). Such costs are not explicit and may be
many times the explicit costs detailed in the wrap-fee contract.160 Lack of attention to trading activities and
to the costs of the adminstrative platforms upon which wealth is managed is the surest way to loose buckets of
money from investment activities.

TAXES, INFLATION AND TURNOVER

For taxable investors, high portfolio turnover increases investments costs because trading activity often
triggers taxable events. Taxable investors must consider the combined impact of trading costs, taxes, and
inflation. These three costs erode returns.
Why arent we all rich? This sentence is the intriguing beginning to a study that appeared in the winter
1995 edition of The Journal of Portfolio Management.161 The authors examined the long term investment
results of several asset classes during the period 1926 through 1993, in order to gauge the effect of taxes,
inflation and trading costs on overall portfolio return.
For trading costs, the authors used commission costs only. To calculate taxes they assumed a single
taxpayer with $75,000 of earned income measured in 1989 dollars and adjusted for inflation, both
prospectively through 1993 and retrospectively back to 1926. Additionally, they applied the actual marginal
rates on both capital gains and ordinary income from 1926 (in 1926 income taxes were 1%, and capital gains
taxes were 6%) through 1993. They assumed 20% portfolio turnover per year. Finally, they inflation-adjusted
the data to determine how much real after-tax, after-trading expense purchasing power investors realized per
dollar invested.
Their findings are displayed in the following graph:

Schwartz & Francioni, Op. Cit, pp. 140-141.

Siegel, S. B., & Montgomery, David, Stocks, Bonds, and Bills after Taxes and Inflation, The Journal of Portfolio
Management (Winter, 1995) pp. 17-25.
160
161

89

Figure 6.3: Ending Value of a $1 for a Variety of Securities

1926 through 1993
$800.08

Nominal
Net of Taxes & Trading Costs
Net of Taxes, Trading Costs, &
Inflation

$157.72

$19.39

S&P 500

$28.03

$10.68

$1.31

Long Term U.S. Gov't Bonds

$11.73

$3.99

$0.49

U.S. T-Bills

S o urc e: S iegel, S . B ., & M ont go mery, Davi d, S to cks, B onds, and B il ls af t er Taxes and Inf l ati on,
T he Jo urnal o f P o rt f o lio M anagem ent (Wint er, 1995) pp. 17-25.

These findings are indeed sobering. Although these tax cost calculations assume a 20% per year portfolio
turnover (i.e., the average security is retained in the portfolio for five years), by mutual fund industry standards,
20% is a low rate of turnover. A query of the Morningstar mutual fund database as of December 31, 2006
reveals that the average actively managed mutual fund turnover rate for the category most comparable to the
S&P 500 - U.S. Large Company Blend Stock Funds is 76%.
What, then, is the relationship between active trading and tax costs? The longer the holding period of the
average security, the longer the tax event of a sale can be postponed. The longer the gains remain unrealized,
the more valuable they are, because deferred taxes on unrealized gains compound for the investor instead of
Uncle Sam. 162 The following graph, taken from a 1993 study, illustrates how various turnover rates influence
ending pre-liquidation wealth for a portfolio operating over a twenty-year period. It assumes an investor with a
combined federal and state capital gains tax rate of 35% and a growth rate of 6%:

Koontz, Warren N. Understanding the Tax Constraints on Private Clients, Investment Counsel for Private Clients (AIMR,
1993) pp. 65-71.
162

90

Figure 5.4: Tax Effect of Turnover:

E ndi ng V alue of a $100 P ortfol io Held for 20 Years
$321

$284

$263

$235

0%

5%

10%

25%

Turnover Rate
S ource: Ko ont z, Warren N. Underst anding t he T ax Co nst rai nts o n P rivat e Cl ient s, Invest m ent C ounsel f o r P rivat e Client s (A IM R, 1993) pp. 65-71

Moving from zero to five percent turnover decreases the effective holding period from 100 years to 20
years (a factor of 5); while moving from 50 percent to 55 percent turnover rate decreases the holding period
from two years to 1.8 years (a factor of 1.1). By the time you reach fifty- percent turnover, most of the tax
damage has already been done. Specifically:

A 5% turnover rate equates to a 0.64% reduction in annual returns;

A 10% turnover rate equates to a 1.05% reduction in annual returns;

A 25% turnover rate equates to a 1.63% reduction in annual returns;

A 50% turnover rate equates to a 1.93% reduction in annual returns.

All else equal, passive (low turnover) portfolios have considerable advantages for taxable investors.

91

C O N C LU S I O N : I N D E P E N D E N T I N V E ST M E N T C O U N S E L

SCHULTZ COLLINS LAWSON CHAMBERS, INC. is an independent firm providing investment counsel to individuals
and institutions. Our business is helping clients through the many obstacles they face as investors. All our
work is based on the premise that our clients are best served, not by attempts to predict the future direction of
markets or prices, but by the practical application of the body of scientific knowledge discovered through
research in financial economics. Our focus, therefore, is on staying current with the research, communicating
its findings to clients, and using it to improve the operation and performance of their portfolios. Indeed, this
document is an important step in service of the latter two goals. It is intended to provide clients with a
theoretical context and grounding for the methods and language we use to construct sound investment policy,
and to constitute a defensible rationale for their asset allocation and portfolio management decisions.
The freedom to apply academic research in our clients behalf comes only with complete independence.
We believe that working in the best interest of our clients precludes any other interest. In particular, we avoid
conflicts of interest by shunning relationships with any firms which would oblige or otherwise motivate us to do
business with them on behalf of our clients, other than by virtue of their financial soundness and the quality of
their products or services.
Our firm is, therefore, not affiliated with any vendor of investment products or services, nor do we offer
advice with the expectation of receiving commissions, finders fees or other similar forms of remuneration.
Rather, we are compensated by clients on a fee basis, with fees calculated as a percentage of assets or at an
hourly rate. Our success is tied to the financial success of our clients. We do not sell investment products. We
provide investment counsel.
SCHULTZ COLLINS LAWSON CHAMBERS, INC. IS an investment advisor registered with and regulated by the
Securities and Exchange Commission, under the Investment Advisors Act of 1940. The Act states that an
investment advisor may be designated as investment counsel only if a substantial part of its business
consists of rendering investment supervisory services.163 The Act defines investment supervisory services as
providing continuing investment advice based on the individual needs of each client.164
SCHULTZ COLLINS LAWSON CHAMBERS, INC. acts as investment counsel. Recognizing that each investor faces a
unique set of objectives and circumstances, we provide unbiased and relevant information regarding the
structure of investment problems and opportunities, the tradeoffs inherent in investment alternatives, and the
range of possible future outcomes. We do not maintain proprietary portfolios into which our clients must fit.
Rather, we design and supervise each portfolio in a manner specific to the objectives and economic
circumstances of each client. We provide information to help you:
Design a portfolio specific to your objectives and economic circumstances;
Improve your chances of realizing a successful outcome; and,
Understand and quantify both the likelihood and magnitude of possible unsuccessful
outcomes.
Our primary objective is to help clients make informed investment decisions. To that end, we are
committed to providing investors and fiduciaries with the information and knowledge that make such decisions
possible. We employ the quantitative and statistical methodology that generally characterizes the science of
decision analysis. In the words of Robert Clemen of the Fuqua School of Business at Duke University:
163
164

15 U.S.C. 80b-8(c).
15 U.S.C. 80(b)-2(a)(13).

92

Instead of providing solutions, decision analysis is perhaps best thought of as simply an

information source, providing insight about the situation, uncertainty, objectives, and tradeoffs, and ... yielding a recommended course of action.165

INVESTMENT POLICY AND THE PRUDENT INVESTOR RULE

SCHULTZ COLLINS LAWSON CHAMBERS, INC. helps investors define their investment policy in a written
Investment Policy Statement. Although an Investment Policy Statement can take many forms, we believe that
all investors, not just trustees and fiduciaries, should demand and receive investment advice that conforms
definitively to the Prudent Investor Rule.166 We therefore acknowledge the principles of prudence, and assume
the requisite duties:

Decisions concerning individual assets must be evaluated in the portfolio context.

Our firm helps investors develop coherent, practical strategies for combining
disparate assets into a viable whole;

Risk and return are directly related. We act on our duty to analyze and quantify risk,
and help our clients make deliberate judgments concerning the level of risk and
return most appropriate for their accounts;

Sound diversification is fundamental to risk management. We assist in minimizing

uncompensated, unwarranted or unintended risk;

The need for current income must be balanced with protection of purchasing power.
We provide clients with the means to achieve real returns on their investments, after
considering taxes and inflation;

A prudently managed portfolio avoids unjustified expenses. Our firm is adamant in its
drive to eliminate unwarranted fees and transaction costs.

OUR APPROACH TO PORTFOLIO SUPERVISION

We use objective mathematical and statistical concepts to help clients build rational, diversified portfolios
that embody the principles underlying the Prudent Investor Rule. Our approach to portfolio management
draws upon favorable characteristics from several portfolio management approaches. Client portfolios are
often diversified globally across a broad spectrum of asset classes. Allocations to each asset class reflect a
careful analysis of each clients risk tolerance and required return.
Each clients portfolio has target allocations (weightings) for the asset classes in which it invests.
Depending on the asset management approach selected by the client, we provide appropriate rebalancing and
trading policy services. For example, for clients electing a Constant Mix asset management approach,
maintaining the weights specified in the strategic asset allocation is critical. Without periodic portfolio
rebalancing to targeted weights, random drift attributable to future performance will create underexposure to
some asset classes and overexposure to others. Nobel Prize winning economist Paul Samuelson describes
this phenomenon in terms of portions (fractions) of the portfolio that inadvertently become over-weighted.
Such bloated fractional representation leads to:

Clemen, Robert, Making Hard Decisions: An Introduction to Decision Analysis, (Duxbury Press, 1996), p. 4.
Restatement of the Law, Third, of Trusts - Prudent Investor Rule, Ch. 7 p. 7
Uniform Prudent Investor Act National Conference of Commissioners on Uniform State Laws
Chicago October, 1994. Adopted by the State of California July 5, 1995

165
166

93

an expensive bargain with the devil. In return for hope of a somewhat higher after-tax mean
total return, you are riskingexcess volatility and risk167
Uncorrected random drift of asset class returns precludes adherence to prior conscious decisions
regarding the appropriate level of risk. Since risk can only be judged within the portfolio context, shifting
portfolio positions make it difficult to manage portfolio volatility.
Our approach is summarized in the chart on the following page.

167

Samuelson, Op. Cit., p. 21.

94

PORTFOLIO MANAGEMENT APPROACH

DETERMINING CLIENT OBJECTIVES

Detailed fact finding to build unique client profile

Hands on assistance to assess individual financial needs and risk tolerance
Investment goals and approach codified in a customized Investment Policy Statement
(IPS)

PORTFOLIO CONSTRUCTION

Asset classes selected and weighted based on required return and acceptable risk
Customized portfolios developed to meet individual needs
Controlled use of actively managed funds
Broad global asset class diversification
Diversification within asset class, mitigating systematic risk
Flexibility to weight specific factors, to enhance return or manage volatility
Turnover managed using passive/indexed strategies

TAX AND EXPENSE MANAGEMENT

No-load investments used to mitigate investment costs

Institutional funds used to minimize fund expense ratios
Tax liabilities managed using low turnover strategies, exchange traded funds
Tax arbitrage, using IRAs and qualified plans to hold investments that otherwise
generate high tax liabilities
Option to use tax advantaged fixed income (muni bond) funds in taxable accounts

PORTFOLIO MONITORING AND SURVEILLANCE

Objective criteria for determining ongoing suitability of specific portfolio investments

Comprehensive portfolio performance reporting
Annual review of portfolio results relative to IPS objectives and guidelines
Current research findings incorporated into asset management process
PORTFOLIO MANAGEMENT

Periodic rebalancing to asset allocation targets

Institutional trading to reduce transaction expenses
Annual rebalancing activity, to avoid impairing portfolio returns in trending markets

95

In addition to assisting with portfolio design and implementation, we provide portfolio supervision services so
that:

Performance results are presented clearly;

Returns are compared to wealth accumulation targets;

Suitability is monitored;

Investment Policy adapts to changing circumstances and current academic research;

Portfolio performance is tracked and evaluated relative to client objectives.

We evaluate current arrangements and provide unbiased recommendations for individuals, trusts,
endowments and retirement plans. Our firm evaluates most areas of trust and plan operations, and has the
expertise to address problems or issues normally considered outside the scope of an investment advisors
responsibilities. With our depth of experience and unbiased perspective, we provide clients with practical,
economical and reliable solutions to the broad array of issues encountered by fiduciaries, plan administrators
and individual investors. We do not provide legal, accounting or tax advice.
Our goal is to lead our clients to a position of understanding. Decisions can then be taken, not on the basis
of hunches, or blind trust in anothers expertise, but in the context of a rational, comprehensible, legally
defensible and academically supported framework.

96