GI Book 6e-Trang-2

Structuring the Global
Investment Process
LEARNING OUTCOMES
After completing this chapter, you will be able to do the following:
■ Discuss the functions of and relation- ment policy statement and strategic
ships among the major participants or tactical asset allocation, of fund/
in the global investment industry manager choices available to a global
investor
■ Discuss the components of a formal
investment policy statement (IPS) ■ Discuss the important issues—par-
ticularly scope, weights, and cur-
■ Evaluate the appropriateness of an
rency allocation—in choosing a
investment policy statement for a
global benchmark for strategic asset
global investor
allocation
■ Discuss the role of capital market ■ Compare and contrast alternative
expectations in strategic and tactical
approaches to hedging currency risk
asset allocation
in strategic asset allocation
■ Interpret capital market data and ■ Discuss the determinants of effec-
capital market expectations in the
tive global tactical asset allocation
context of strategic asset allocation
for a global investor ■ Compare and contrast strategic and
tactical asset allocation in the global
■ Compare and contrast the major
context
choices—active/passive, top-down/
bottom-up, style, global/specialized, ■ Describe and evaluate the compo-
currency, quantitative/subjective— nents of the global asset allocation
available to a global investor in struc- process
turing the global investment deci- ■ Design an appropriate strategic
sion-making process
asset allocation for a global investor
■ Evaluate the appropriateness, in
terms of implementing the invest-
From Chapter 13 of Global Investments, 6/e. Bruno Solnik. Dennis McLeavey. Copyright © 2009 by Pearson Prentice Hall.
All rights reserved.
527
Structuring the Global Investment Process
■ Interpret capital market data and ■ Evaluate the implications of a port-

capital market expectations in the folio performance analysis for a
context of tactical asset allocation global investor
for a global investor ■ Summarize the primary issues that
■ Design an appropriate tactical asset must be addressed by an investor
allocation for a global investor contemplating global investment
A n investment organization that is planning to offer global investment products

to its clients first needs information on institutions and techniques. Then the
manager needs to properly structure the entire global investment decision process,
from research to management and control. Among the many global investment
organizations in the world today, there are a great variety of approaches that reflect
different investment philosophies and strategies.
This chapter is organized as follows. The first section is a brief tour of the
global investing industry. The second section reviews the various approaches to
choosing a global investment philosophy to propose to a client. The rest of the
chapter discusses the various steps of the global portfolio management process. As
discussed in the third section, the client’s objectives, constraints, and requirements
are specified in an investment policy statement that forms the basis for the strategic
asset allocation. Finally, capital market expectations are formulated for use in
determining the asset allocation. This crucial step in a world with many equity,
debt, and currency markets is discussed in the fourth section. Some important
issues in terms of global asset allocation are then discussed. Strategic asset alloca-
tion and tactical asset allocation are usually viewed as two different steps of the asset
allocation process. The chapter concludes with an example of how to structure and
quantify the global investment process.
A Tour of the Global Investment Industry
Readers unfamiliar with the global investment scene often ask for a brief presentation
of the major players. There are several types of participants in the global investment
arena: investors (private or institutional), investment managers, brokers, consultants
and advisers, and custodians. Some players belong to several categories.
Investors
Private Investors Investors belong to two broad categories: private and
institutional. The term private investor usually refers to an individual or a small
group of individuals, such as those represented by a family trust. Private investors
have many ways to invest globally. They can buy foreign shares listed on their
domestic markets. They can also buy, through a broker, foreign shares not listed
528
domestically. They can buy mutual funds that diversify across multiple international
markets or funds that invest only in specific markets. Finally, they can have their
money managed by investment professionals; high-net-worth individuals are a
clientele actively sought by many investment management firms. Whether a private
or institutional investor, the client should prepare, with the manager, a policy
statement outlining the investment goals and the risks that can be taken. This
procedure is all the more important in global investing, in which the investment
universe is enormous, as are the potential returns, costs, and risks.
Institutional Investors The term institutional investor is generally used to

describe an organization that invests on behalf of others, such as a mutual fund,
pension fund, or charitable organization or an insurance company. A variety of
institutional investors exist. Mutual funds pool money from investors and invest it in
financial assets to achieve a stated investment objective. Mutual funds are offered to
the public at a listed price reflecting the market value of the fund’s assets. Mutual
funds are called unit trusts in the United Kingdom or Sociétés d’Investissement à
Capital Variable (SICAV ) in France, Belgium, and Luxembourg.
Retirement provisions are diverse throughout the world and even within each
country. Basically, two very different philosophies can be found. Many countries
have a pay-as-you-go system: Active workers pay for the pensions of retired workers.
Workers, and their employers, do not contribute to the future pensions of the
employees; they do not “capitalize” their contributions. Rather, these contributions
are immediately paid to the current retirees. Except for some minimal technical
reserves, there is no money to be invested, because it is immediately disbursed.
In a capitalized contribution system, employers contribute to a pension fund, which
capitalizes all contributions and pays them back at the time of retirement. In this
case, pension funds are considered long-term institutional investors. In general, a
basic social security (pay-as-you-go) system tends to coexist with supplementary
pension funds (sometimes called second pillar ).
The investment approach of pension funds is greatly affected by the way future
benefits are planned. There are basically two types and a combination thereof. A
defined benefit (DB) pension plan promises to pay beneficiaries a defined income after
retirement. This benefit depends on factors such as the workers’ salary and years of
service. The contributions are based on an actuarial rate of return assumption; capi-
talized at this actuarial rate, the contributions should be equal to the promised bene-
fits. If the investment performance of the pension assets is insufficient, over time, to
meet this assumption, the employer may have to increase its contribution to cover
the shortfall. Therefore, the corporation carries the risk of the pension plan. This
type of plan leads to investment policies that focus on meeting the return objective
over the long term while reducing the risk of a major shortfall in the near term.
Defined contribution (DC) pension plans are increasingly favored worldwide. In a
defined contribution plan, the amount of contributions is set, usually as a percent-
age of wages, but future benefits are not fixed. If the rate of return on invested
assets is high, large benefits will be paid at time of retirement. If the rate of return
is low, smaller benefits will be paid.
529
For either type of plan, a board selects a long-term asset allocation. Investment
management is either delegated to external managers, done in-house, or a combina-
tion of both. The selection of investment managers is a lengthy process, often assisted
by consultants. External managers are given a mandate, detailed in an investment
brief. For example, one manager could be selected to manage a portfolio invested in
emerging stock markets. The mandate will state the investment objective (e.g.,
“Outperform the S&P/IFC composite index”) and provide guidelines on the type of
investments that are acceptable and the amount of risk that can be taken.
In a traditional pension fund, all contributions are pooled and the total money
is managed collectively. A recent trend is to give more investment decision power to
each employee. The plan sponsor offers a range of asset allocations and investment
products from which each employee can select. Hence, external investment man-
agers have to pass two selection hurdles: the plan sponsor and the individual
investors. Pension reforms in many countries follow the U.S. model and give
employees some flexibility in asset allocation.
Endowments and foundations accumulate the contributions made to charitable
and educational institutions. Endowments are long-term investment portfolios
formed from donated funds and restricted to maintaining their principal over
time. Foundations are entities that provide grants to accomplish the purposes of
the original donor or donors. They have a variety of horizons (some are designed
to terminate after a specific number of years) and spending rates (some have
legally determined minimum spending rates), but most foundations have in com-
mon the need to meet their spending entirely from investment returns.
Endowments and foundations tend to have great investment freedom because
they operate under fewer regulatory constraints than other institutional investors.
As a result, their return objective tends to focus on total return over the long run.
Their investment policies are often the most aggressive among institutional
investors, with many having extensive global and alternative investments.
Insurance companies form another major category of institutional investors. Life
insurance companies collect premiums for insuring lives, and these premiums are
invested until claims are paid. Insurance companies are heavily regulated in each
country and state in which they operate. They tend to adopt conservative investment
policies, focusing on fixed-income assets, in order to assure their claim-paying ability.
Global asset allocation is often limited by regulatory constraints. Property and casualty
insurance companies operate under a different regulatory framework that is some-
what more permissive in all countries. The investment policy of any insurance com-
pany depends on its surplus. Loosely speaking, the surplus is the difference between
the total value of the invested assets and the expected claim payments. Although the
assets matching the expected claim payments are typically invested in domestic bonds,
the surplus can be invested in riskier assets, including global investments.
Investment Managers
Investment managers form a diverse group. It is hard to offer a single classification
of asset managers. They range from the asset management department of banks to
530
independent boutiques specializing in offering specific investment products.

Some managers offer a diversified menu of products (much like a supermarket)
and basically charge a fee based on the assets under management, while others
offer only a specialized product promising high investment performance and
may charge fees linked to their performance. Some asset managers cater to retail
as well as institutional clients, while others, sometimes less known by the
public, serve the needs of only one type of client, such as a pension fund. Others,
such as hedge funds, are a highly specialized type of asset manager. Many asset
managers now offer global investment capabilities, either directly or through joint
ventures with foreign partners. U.S. asset managers that offer global investment
management to U.S. pension funds can easily offer a similar product to British,
Dutch, or Swiss pension funds, so competition among managers has truly become
global.
Brokers
Stockbrokers, or brokers, have traditionally played an important role in asset
management, both in terms of implementing security trades and in research of
companies and markets. They have a large staff of financial analysts who cover
companies domestically and sometimes worldwide. This financial analysis is called sell
side because brokers use this research to sell trading ideas to investors (buy side analysts
work for investment managers and institutional investors). In many countries, the
standards of professional conduct are lax compared with those in the United States or
the United Kingdom. However, all CFA charterholders and CFA candidates must
follow the CFA Institute Code of Ethics and Standards of Professional Conduct,
wherever they work and invest. Many brokers also act as investment banks, but the
independence of sell-side analysts must be maintained.
Major global brokers have offices in all major financial centers and cover most
markets in the world. Other brokers specialize in some niches (e.g., European small-
cap stocks). Bond brokers provide extensive research on interest rates, exchange
rates, and company and country risk. The research efforts at these brokerage houses
have moved in two directions. As a result of their excellent research capabilities, bro-
kers moved successfully into investment banking. Financial analysts extended their
approach to cover mergers and acquisitions as well as corporate finance. Others
applied their research expertise and knowledge of the financial markets to asset man-
agement. Asset management is a low-margin business compared with investment
banking, but profits are much more stable. Brokers in Japan have always offered asset
management, and this service is now a major priority for the big U.S. brokerage
houses, which are marketing their asset management skills in Europe and Asia.
Consultants and Advisers

Consultants and advisers provide several types of advice and play a major role in the
asset management industry. Although they are better known for their work with the
pension fund market, they are also employed by other investors, such as insurance
531
companies, endowments, foundations, and high-net-worth individuals. Independent

consulting firms have traditionally advised U.S. pension funds, while actuaries played
a similar role in the United Kingdom. For many years, consultants played no role in
most other countries, but this situation changed dramatically in the late 1990s.
Consultants can help with all the actuarial calculations of pension plan obliga-
tions, such as liabilities to plan members and technical reserves. These calculations
depend on many factors, such as the age structure of plan members, retirement
patterns, the plan’s benefits formula, and so on. Although these actuarial calcula-
tions are often performed in-house in the United States and other countries, the
United Kingdom traditionally requires the intervention of external actuaries.
Consultants also focus on services such as recommending asset allocation, selecting
investment managers, and monitoring performance, and they may give tax and
legal advice.
Major consultants measure the performance of a large number of managers
with different mandates. Hence, they maintain a database that allows them to
compare the performance of a universe of managers. But probably the most sensi-
tive role of consultants, which has spread throughout Europe and the United
States, is their assistance in the process of selecting, hiring, and firing external
managers. Some commentators state that pension fund sponsors are relieved to
share their fiduciary responsibilities with consultants who have a worldwide repu-
tation and are used by numerous other pension funds. When looking for an exter-
nal manager, a pension fund typically prepares a request for proposal (RFP ) or request
for information (RFI ), which specifies the type of manager sought. The consultant
organizes the selection process by sending out the RFP or RFI in the form of a
questionnaire to be completed by prospective managers. Based on the responses
and the consultant’s own research on the managers, a short list is established, and
managers on this list make a presentation to the pension fund, detailing their
case. The board of the pension fund decides, with the consultant’s advice, which
manager(s) will be selected. It should be noted that the arrangement between
pension plan sponsors and consultants varies considerably. Some plan sponsors
delegate much of the decision-making process regarding manager hiring and fir-
ing to the consultant, while others prefer to make the decision themselves and use
the consultant in an advisory role. Many consulting firms advise only domestic
institutional clients. But some consulting firms have developed a client base in
many countries, either by growing globally or by participating in a global network
of consultants.
A new trend for pension consultants is to go one step further and offer multi-
manager funds or funds of funds (FoFs). The consultant does not manage money
directly but creates a fund of selected managers. This approach has caused some
controversy over a potential conflict of interest in the advisory capacity.
Individual investors are also in need of professional advice for the manage-
ment of personal or family assets. As discussed earlier, the taxable environment of
private investors poses a challenge, and tax systems vary markedly across countries,
even within the same region. For example, the tax environment is different for
each country in the European Union. Hence, advisers tend to be domestic rather
than global players, but they may be part of a global network.
532
Custodians
Securities owned by investors are deposited with a custodian, which often uses a
global network of subcustodians. Given the complexities of global investing due to
national differences in taxation, trading, and settlement procedures, an efficient
and reliable custodian is a necessity. Accurate information from around the world
should be gathered rapidly. Automated trade notification is necessary. Computer-to-
computer links with the subcustodians, clearing services, and the manager should
be set up. Income collection should be swift and correctly reported. Tax recovery
should be automated and carefully checked with each government. Securities-
lending facilities should be available in most markets. Finally, a cash management
system in many currencies should be implemented. Cost and reliability are two
important features of custodial services.
Information technology is an important component of custodial services. With
the high development costs of such software, many banks have sold their custodial
activities, and further consolidation is expected in the future because economies of
scale can be significant in this business.
Global Investment Philosophies
Global investment has grown rapidly among institutional investors of all nationalities.
Investment managers worldwide can no longer treat foreign investments as an exotic,
minor consideration, and they must now have a clear view of how to approach global
investing. An investment management organization must make certain major choices
in structuring its global decision process, based on several factors:
■ Its view of the world regarding security price behavior
■ Its strengths, in terms of research and management
■ Cost aspects
■ Its location and prospective domestic/global marketing strategy
Similarly, investors will select managers based on the same factors, as well as their
ability to control currency risks. These important choices are discussed in the follow-
ing sections in terms of investment philosophy and strategy from the viewpoint of a
fund manager. But individual or institutional investors face exactly the same choices
when deciding on their approach to global investing and their selection of fund
managers. The global approach to equity is more complex than the approach to
bonds, so some of the points discussed next are primarily relevant for equity portfo-
lio management.
The Passive Approach

A fund managed according to a passive approach simply attempts to reproduce a
market index of securities. This type of fund is often called an index fund. The sole
533
purpose of an index fund is to exactly track the return on a selected market index,
to capitalize on its long-term performance while keeping all costs at a minimum.
The passive approach is an extension of modern portfolio theory, which claims that
the market portfolio should be efficient. In the United States, the domestic index
fund approach is supported by extensive empirical evidence of the efficiency of
the stock market. A similar domestic index fund approach has developed in the
United Kingdom and, more recently, in other European countries and Japan.
Large pension funds have moved extensively to indexing their domestic assets.
The trend toward global indexing is strongly felt among institutional investors.
It is now common to see funds that passively replicate some international index
(e.g., some Europe or ex U.S. benchmark) or a World index. While institutional
investors can invest in dedicated passive portfolios, retail investors can choose from
some passive mutual funds, including exchange traded funds (ETFs) that track
some international index.
The basic argument supporting the passive indexing approach is that the alter-
native, an active strategy, requires above-average ability in forecasting markets, cur-
rencies, and security valuation. Forecasting is never easy; moreover, it entails higher
commissions and costs. With a passive strategy, the fund can achieve the full bene-
fits of global risk diversification without incurring these high costs.
Various indexing methods can be used:
■ Full replication: All securities in the index are bought, with proper weighting.
This method is impossible globally, given the huge number of securities
involved.
■ Stratified sampling: This approach tracks the index by holding a representative
sample of securities. The securities are grouped according to various criteria
(country, firm’s size, industry, yield), and the index fund sample attempts to
replicate the characteristics of the index across the various criteria.
■ Optimization sampling: This is a sampling method using factor models to mini-
mize the tracking error of the index. It is a sophisticated statistical method
based on a large number of factors, or firm attributes, and using optimiza-
tion techniques based on historical relationships.
■ Synthetic replication: A stock index can be replicated by using a futures con-
tract on the index plus a cash position. The fair pricing of the futures
ensures that the index is closely tracked and that transaction costs are low.
For some investors, legal aspects regarding the use of derivatives constrain
the implementation of this approach on a global scale. Another problem is
that futures contracts tend to be written on indexes based on a subset of the
broad market, while natural market benchmarks are broadly based indexes.
Stock index swaps are also used.
Although indexing does reduce costs, perfect tracking of a global index is not feasi-
ble. For example, whenever a dividend is paid in one currency—say, the Australian
dollar—it should be reinvested simultaneously in all global securities with proper
534
weights, which is an impossible task. Indexing requires superior computer, adminis-

trative, and trading technology, but it does result in large economies of scale.
Because the quality of the indexer shows up quickly, and objectively, in the form of
a low tracking error, only the best indexers are likely to capture a large market
share of this management style. A remaining issue, discussed later, is the choice of
the global benchmark to be passively tracked.
The Active Approach

In an active strategy, investors place bets on the various factors that affect securities’
behaviors. A benchmark is often imposed in the mandate set by the client, and it
will clearly guide the structure of the portfolio. Active deviations from the
benchmark can lead to superior performance. Of course, the strategy itself
depends on the investor’s view of the world. Active decisions of a portfolio manager
show up at various levels:
■ Regional/country allocation: This is the choice of national markets and curren-
cies. The manager can select long-term weights that differ from those of mar-
ket indexes, such as the EAFE index. This long-term allocation is often called
strategic asset allocation. Periodic, temporary revisions in the weights can be
justified by changes in market expectations or risk estimates. This is often
called tactical asset allocation. Allocation decisions also are discussed later in
the chapter.
■ Sector/industry selection: The globalization of the economy pushes some man-
agers to allocate funds across worldwide industries rather than countries.
This is more commonly done when managing a portfolio within integrated
regions such as Continental Europe.
■ Style selection: Some managers believe that companies with similar attributes
tend to have similar stock price behavior. Common style decisions are value
versus growth stocks or small versus large firms.
■ Security selection: To achieve a given asset allocation, managers can engage in
active selection of securities within each asset class and conduct pair arbi-
trage between two similar securities.
■ Market timing: Managers can resort to market timing to temporarily increase
or reduce the exposure in one or more markets or currencies. This is a short-
term trading tactic (as opposed to long-run strategies for allocating assets)
that often involves the use of derivatives.
■ Currency hedging: Managers can actively manage their currency position.
A few additional dimensions can be found in the active management of global debt
securities:
■ Sector selection/credit selection: Managers can deviate from a bond benchmark
by overweighting some market sectors or investing in higher credit risk, such
535
as low-credit-rating bonds and emerging debt. This strategy leads to an

increase in yield but also in credit risk.
■ Duration/yield curve management: In each currency market, the manager can
adjust the duration of the portfolio according to forecasts about changes in
the level of interest rates and shapes of the yield curve.
■ Yield enhancement techniques: Numerous techniques are proposed to add value
to the performance of the basic strategy; these include lending securities,
using swaps and other derivatives, investing in complex bonds, and doing
pairs arbitrage.
A manager can be active in some dimensions and not in others. For example, active
asset allocation management can be achieved using national index funds for each
market. Conversely, a manager could follow a fixed asset allocation, using active
security selection within each market.
The risk diversification argument is usually at the heart of the rationale of both
active and passive global money managers. In the quickly changing global environ-
ment, active money managers must be able to evaluate and control the risk of their
portfolios as rapidly as possible. Taking active bets on markets, currencies, or secu-
rities could result in portfolios that turn out to be quite risky and do not provide
the global risk diversification benefits so widely claimed.
Following is a review of major choices to be made in active global asset manage-
ment. It focuses primarily on equity because the management of portfolios of debt
securities is more technical.
Balanced or Specialized
Investment managers for private clients have traditionally been balanced money
managers. For example, a European bank advising a private client will suggest both
the global asset allocation and the selection of each security for the portfolio; the
bank will advise on all aspects of private wealth management. The asset allocation
covers all asset classes: equity, debt securities, alternative investments, and cash, as
well the currency hedging policy.
Even for pension assets, the approach of having a single balanced manager or,
at most, a few can be used. In this case, the investment policy statement can be
quite loose, with objective statements such as “Maximize long-term returns subject
to an acceptable level of risk normally associated with a balanced approach to fund
management.”1
By contrast, the specialization approach involves hiring managers who special-
ize in particular investment areas, such as Japanese stocks, European growth stocks,
or emerging market debt. The trend toward specialized global management is
based on the hypothesis that no manager is an expert on all markets, although
some managers can be superior on one or a few markets.
1
Extract from Mark Tapley, “Lessons from the Unilever/Merrill Case,” IPE, May 2002, p. 36.
536
For a client, the central question in the specialized approach is how to decide
on asset allocation. Because asset allocation is potentially the most profitable and
important investment management decision, it seems a bit odd to deprive asset
managers of this important opportunity to add value. This specialization constraint
is especially problematic in equity management when listed companies compete
globally and stock markets are increasingly correlated. To study European equity,
one must compare European firms with their U.S. and Asian competitors. If an
equity manager is expert on European equity, she should (must) also be expert on
global equity.
Industry or Country Approach

Asset managers should structure their investment processes along the major factors
affecting security prices:
■ If a manager believes that the nationality of a company is a major factor
affecting the return and risk on the company’s stock, then the global portfo-
lio should be structured primarily according to country or region. The allo-
cation to countries/regions is a major investment decision; then stocks are
selected within countries. Although there is a tendency to presume that the
country of origin or the head office of a multinational firm is irrelevant, a
close look at human resource and hiring/firing constraints will reveal that
such an idealized world is not yet here.
■ If a manager believes that the industry or sector in which a company operates
is a major factor affecting its return and risk, then the global portfolio should
be structured primarily according to industry or sector. The allocation to
industries/sectors is a major investment decision, and then stocks are
selected within global industries.
■ If a manager believes that the major factors affecting the return and risk on
the company’s stock are specific to that company (quality of management,
cost structure relative to competition, prospects for its products, etc.), the
manager should rely primarily on company analysis.
Top-Down or Bottom-Up
Both domestically and globally, portfolio managers use either a top-down or a
bottom-up approach or combination to the investment process.
■ Top-down approach: The manager using a top-down approach first allocates his
assets across asset classes and then selects individual securities to satisfy that
allocation. The most important decision in this approach is the choice of
markets and currencies. In other words, the global money manager must
choose from among several markets (stocks, bonds, or cash) as well as a vari-
ety of currencies. Once these choices have been made, the manager selects
the best securities available in each market. In global equity portfolios,
537
a manager following the country approach decides on the percentage allo-

cated to the various countries and then selects the best stocks in each market;
a manager following the industry approach decides on the percentage
allocated to the various global industries and then selects the best stocks in
each industry.
■ Bottom-up approach: The manager using a bottom-up investing approach studies
the fundamentals of many individual stocks, from which she selects the best
securities (regardless of their national origin or currency denomination) to
build a portfolio. For example, a manager may be bullish on car manufactur-
ers and buy shares in all of them (GM, Toyota, Volkswagen, Peugeot, and oth-
ers), or the manager may buy shares in only the best oil company in the
world, regardless of its national origin. The product of this approach is a
portfolio with a market and currency allocation that is more or less the ran-
dom result of the securities selected. Implicit in this approach is the man-
ager’s greater concern with risk exposure in various sectors than with either
market or currency risk exposure. But currency risk or market exposure can
be hedged.
■ Top-down/bottom-up approach: Some investment organizations attempt to com-
bine a top-down with a bottom-up approach, but this is not easy to achieve.
The rationale for combining the two often cites risk control. For example, a
top-down manager may choose to limit exposure to a particular country
when the bottom-up manager would decide otherwise. The entire invest-
ment decision process has to be consistent, and sophisticated risk manage-
ment tools are needed. A manager typically bases his security selection on
worldwide company analysis. Factor models are used to control the exposure
to countries and industries. Pair trades between companies in the same
industry (but different countries) or in the same country (but different
industries) allow the manager to satisfy risk objectives in terms of geographic
and sector/industry allocation, while taking advantage of superior company
analysis (see Example 1).
Style Management
Some style managers construct portfolios based on some attributes of
companies. This approach has been extended to global investing. Common style
decisions are value versus growth stocks or small versus large firms. In the 1980s
and early 1990s, a simple strategy that favored small and value stocks yielded
above-average returns in the United States. The simplest global extension was to
choose stocks based on the firm’s size and its price-to-book ratio. However, style
analysis does not extend globally in a simple manner (see Michaud, 1999). For
example, a firm of a given size may look small in the United States and rather
big in France. Similarly, a given style indicator, such as the price-to-book ratio,
which works well in identifying a value factor in one country, may not work in
another country, in part due to accounting differences. To define a value stock,
538
EXAMPLE 1 PAIR TRADES
You own a well-diversified global portfolio mimicking the World index. You
wish to retain the same regional and industry exposure as the World index. You
believe that Total, a French oil company, is undervalued (much lower P/E) rel-
ative to ExxonMobil. You also believe that BMW is overpriced relative to
General Motors. You own $100,000 of ExxonMobil shares and $120,000 of
BMW shares. What could you do?
SOLUTION
■ Sell the shares of ExxonMobil in the portfolio and buy $100,000 of Total
shares.
■Sell $100,000 of the BMW shares and buy $100,000 of General Motors
shares.
The net result will leave unchanged the asset allocation to European and
American stocks. It will also leave unchanged the global industry allocation.
managers resort to different indicators in different countries. Careful analysis is

required to derive style factors that would apply to companies in all countries.
Furthermore, a given style factor is not necessarily synchronized across countries.
For example, value stocks could underperform growth stocks in France, while the
reverse could simultaneously be true in Germany. Finally, no style can systematically
outperform others forever. This was amply demonstrated by the poor relative
performance of small stocks and value stocks in the late 1990s. This does not mean
that active style management cannot achieve superior performance. Style rotation
strategies can add value if the manager can forecast when a given style will become
favored by market participants (or lose favor).
Currency
Some managers treat currencies only as residual variables; their currency
breakdowns are determined by the countries, industries, and securities they select
for their portfolios. They generally consider currency risk a necessary evil and
argue that currency movements are impossible to predict and wash out in the long
run anyway, because it is the real economic variables that ultimately determine a
portfolio’s performance.
Other managers fully hedge their foreign portfolios or decide on a permanent
hedge ratio based on a historical estimate. Full hedging is based on the belief that
foreign currency risk premiums are small or unpredictable and, hence, that
investors are not compensated for carrying foreign exchange risk. Others adopt a
passive, uniform currency-hedging strategy with a fixed hedge ratio different
from 100 percent (e.g., 50%).
539
Still other managers take a proactive tactical approach to currency forecasting.

They try to minimize the contribution currency makes to total risk, and cash in on
opportunities created by currency movements.
At the extreme end of the spectrum is a breed of global money managers,
called currency overlay managers, who actively manage the currency exposure of a
portfolio and often resort to currency options and forward or futures contracts for
selective hedging and speculation. Currency overlay managers do not manage the
portfolio itself. The client transfers the composition of the portfolio managed by
another party on a daily basis. This other party makes no currency-hedging deci-
sions but only manages the assets. The currency overlay managers take currency
positions on the portfolio using currency derivatives. Some even offer pure cur-
rency products attempting to generate high returns from currency plays.
Naturally, the approach an investment manager takes toward currencies leads
to very different portfolio strategies. The question of currency hedging is raised
again later in this chapter.
Quantitative or Subjective
Whether a manager employs a quantitative or a subjective decision-making process,
a great deal of information is required before portfolios are constructed.
Quantification of the investment process is very helpful in global investment
management because of the large number of parameters and decision variables
involved. Quantification can be applied to various models or aids used in the global
investment process, including the following:
■ Econometric or technical forecasting models of markets and currencies
■ Global asset allocation optimizers
■ Dividend-discount models, factor models, duration models, or option valua-
tion models (for quantitative assessment of individual securities)
■ Risk management models
■ Performance and risk analysis
Managers who favor the subjective approach argue that the global environment is
too complex to permit formal quantification. They think that models based on past
data are not helpful because constant changes in the global environment distort
such models. Quantitative managers also believe that the global environment is
complex, but they eventually come to the opposite conclusion regarding the help-
fulness of models. Models are used to extract from past data some useful lessons
for the future. An advantage of a purely quantitative approach is that it is highly
disciplined and therefore avoids some emotional reactions that can lead to poor
decisions and impaired performance.
Many managers rely on subjective judgment but make use of some quantitative
models in their portfolio management process. Others tend to follow models but
inject some judgment as they think appropriate.
540
The Investment Policy Statement
An investment policy statement (IPS) is the cornerstone of the portfolio management

process. The IPS is prepared by the advisor and the client and serves as the
governing document for the portfolio manager and for all investment decision
making. The IPS is extensively discussed in Maginn et al. (2007), and we summarize
only its main features and stress aspects that should be taken into account when
investing globally. A case study, used as example throughout this chapter, details
the preparation of such a document. The case circumstances are described in
Example 2.
EXAMPLE 2 JOHN BOUDERI CASE CIRCUMSTANCES
Leigh Brennan is a financial advisor working with John Bouderi.2 Bouderi retired
at age 60 in March, selling the business and receiving net proceeds after tax of
$1.5 million. Because Bouderi is Australian, the dollar symbol in this case refers to
Australian dollars. He is looking forward to a long retirement because his parents
lived a relatively healthy life to an old age. He expects to do the same.
BACKGROUND
Bouderi’s wife died ten years ago. He has two children, both now in their late
twenties and self-sufficient. Although Bouderi has no grandchildren now, he
hopes for them someday and would like to provide a financial legacy for them.
He desires a comfortable retirement and is determined that this objective not
be jeopardized by poor investment decisions. Bouderi has developed an under-
standing of investments, gained primarily from years of personal study and
observing markets, and has begun to develop and implement his own invest-
ment strategy.
Given his long time horizon, Bouderi initially decided to invest $600,000 in
direct purchase of Australian equities. Although he understood the benefits of
global investing, he limited his initial investments to Australian issues such as
AMP, ANZ, and Westpac. This decision was due partially to his familiarity with
the companies, but also because he felt that the Australian market had shown
strong performance. A subsequent 5% drop in the Australian stock market
made Bouderi question his decision to invest solely in Australian shares.
The $900,000 remaining was added to an existing money market account,
bringing the total money market balance to $1 million. The account currently
yields 5 percent and will be used to cover his living expenses while awaiting
further investment.
The first several meetings between Bouderi and Brennan were devoted to
evaluating Bouderi’s current financial circumstances in detail. He provided the
following information on his financial situation.
2
Case developed with Jan Squires, CFA, and Victoria Rati, CFA.
541
Assets
Residence in Sydney, Australia; market value: $1,000,000
Domestic equity portfolio: 570,000
Money market account: 1,000,000
Total: $2,570,000
Estimated after-tax annual income desired: $90,000
Immediate Plans
1. Take a family vacation in Europe: estimated cost $50,000
2. Sell existing home and purchase smaller retirement home for approxi-
mately 520,000.
3. Place net cash proceeds from home transactions in investment portfolio
4. Establish a trust for any future grandchildren with initial contribution of
$100,000
Expected investable assets after completion of immediate plans:
$1,900,000 (= $2,570,000 - $50,000 - $520,000 - $100,000)
Liabilities
Current home is fully paid off; no mortgage or any other loans are out-
standing.
RISK AND RETURN OBJECTIVES

Brennan next conducted several interviews to determine Bouderi’s investment
goals and to identify any constraints that might limit his achieving those goals.
Desired Return
Bouderi desires to lead a comfortable retirement, with an income that
keeps pace with inflation. He prefers steady returns consistent with a mod-
est level of risk and desires a return from equity investments of 10 percent
annually before tax. For his overall portfolio, he decides that 7 to 8 percent
annually before tax would be acceptable.
Required Return
Brennan believes that Bouderi’s required equity return is in excess of
9 percent, given total investment assets, liquidity needs, time horizon, and
inflation protection needs. Bouderi finally agrees with Brennan’s figure,
but he is concerned about the level of risk a higher return target will bring
to the portfolio.
Risk Tolerance
Bouderi’s personal investing style suggests he is prudent, diligent, and
methodical. This is consistent with the approach he took in running his
542
business. Despite his recent experience in the equity market, he still believes
that it is necessary to take calculated risks to realize appropriate rewards.
In their final discussion about risk tolerance, Bouderi indicates his willing-
ness to accept a possible 10 percent decline in the value of his portfolio in any
one year in order to achieve his objectives. Brennan also introduces the con-
cepts of ability and willingness to take risk. She discusses the idea that ability to
take risk is determined by such factors as wealth position, annual spending
requirements, and inflation expectations. Bouderi realizes that the two con-
cepts of ability and willingness may be substantially different but is unsure how
to combine them to understand his own risk position.
IPSs vary in terms of content, especially between private and institutional

clients. IPSs for private clients often reflect a balanced mandate to manage the
whole wealth. But an IPS could be more specialized and target specific asset classes.
For example, an investor could give a mandate to manage international assets and
prepare an IPS accordingly. A well-constructed policy statement typically includes a
summary of the various elements:
1. Client description and purpose: The IPS starts with a brief description of the client
and the purpose of the establishment of the investment policy statement. It could
detail the duties and investment responsibilities of the various parties involved.
2. Return and risk objectives: Establishing objectives for return and risk is an impor-
tant step for individual as well as institutional clients. The manager must be
sure that investment goals, especially the return objective, are attainable within
the client’s risk tolerance. Some institutional investors set quantified objectives.
For example, a British pension plan could set as an objective for a manager to
outperform the World equity index, ex U.K, by 3 percent per year, with a maxi-
mum annual tracking error of 4 percent. Return is measured as total return
from income and capital appreciation, which is specified as real or nominal,
and as pretax or after-tax. Investors have stated return desires that may or may
not be realistic. Nevertheless, the adviser and the client also must determine
the investor’s required return, the return that must be achieved on average.
The required return is necessary, as contrasted with the stated return, which
may be more in the nature of a hope. The return objective must be set to
achieve the investor’s required return balanced with the investor’s risk toler-
ance. Risk tolerance is the capacity to accept risk; risk aversion is the inability and
unwillingness to take risk. Risk tolerance is a function both of the investor’s
ability and his willingness to take risk. An investor may have more willingness to
take risk than he has ability to accept the consequences of large losses. Ability is
a function of such characteristics as income needs and asset base, as well as
liquidity requirements and time horizon. With basic statistics and the simplest
543
assumption of returns being independent and normally distributed, the sam-

ple mean and standard deviation can be used to explore the probability of
losses for the investor. If these are found to be tolerable, the investor’s risk
tolerance may need calibration.
3. Constraints: All economic and operational constraints on the investment portfo-
lio should be identified in the IPS. Portfolio constraints generally fall into one
of five categories:
■ Liquidity requirements stem from liquidity events when the investor must
make specific cash payments higher than normal long-term net cash flows.
An example of a liquidity event is the planned purchase of a house in one
year. Even though stocks are liquid, their future price is uncertain. The
liquidity requirement constrains asset allocation because the portfolio should
include risk-free government securities targeted to supply the cash required.
As a practical matter, the presence of both liquidity risk (the need to sell less
marketable assets) and price risk (the risk of fluctuations in market prices)
means that the investor must anticipate liquidity requirements by holding
some proportion of assets that are both liquid and relatively low-risk.
■ Time horizon is the period associated with the investor’s objectives. Long-
term horizons are generally considered to be those over ten years, but many
investor horizons are multistage horizons, mixtures of short- and long-term
objective horizons. A long-term retirement objective and a short-term col-
lege financing objective would produce a multistage time horizon. Because
different investments are appropriate for different horizons, the investor’s
time horizon constrains asset allocation.
■ Tax concerns constrain asset allocation decisions because taxable investors
must look for after-tax returns in their taxable portfolios, as well as plan for
tax payments on retirement distributions in their tax-deferred portfolios.
■ Legal and regulatory factors constrain asset allocation because they may rule
out certain investments. In some countries, pension funds are limited to
certain asset classes. Legal and regulatory factors are constraints external to
the investor and imposed by others.
■ Unique circumstances constrain asset allocation. They include constraining
factors other than liquidity requirements, time horizon, tax concerns, and
legal and regulatory factors. These other constraints are internal to the
investor and limit his choice of investments or asset classes. Investor prefer-
ences, investment knowledge, staff, and capabilities are examples of unique
circumstances. Typical constraints are illustrated in Example 3.
4. Asset allocation considerations: The IPS can state some asset allocation objectives
and constraints that have to be followed by the manager. These can be varied,
and a few examples are listed. In particular, the IPS could set out the following
points:
544
EXAMPLE 3 JOHN BOUDERI INVESTMENT CONSTRAINTS
With agreement on investment objectives, Brennan and Bouderi must consider

the constraints on portfolio choice and the attainment of objectives.
TIME HORIZON
Based on Bouderi’s family history of longevity and his current good health,
Brennan determines that he can be expected to experience another twenty
years of active retirement.
LIQUIDITY
Apart from funding the trip to Europe and establishment of the trust, Bouderi
has no unusual liquidity requirements.
TAX ISSUES
All income earned by Bouderi, regardless of source, is assumed to be taxed at a
15 percent annual rate.
LEGAL AND REGULATORY ISSUES

Neither Bouderi nor Brennan has identified any material legal or regulatory
issues facing Bouderi, including the establishment of the trust.
UNIQUE CIRCUMSTANCES
■ The anticipated exchange of homes within the next year will increase
the size of Bouderi’s portfolio substantially.
■ The $100,000 to be set up as a trust for his future grandchildren
would establish an investment account to manage the funds on behalf
of the future beneficiaries. A share of the funds would be available to
each grandchild at age 21. Should there be no grandchildren, the
funds would be available to Bouderi’s children for their retirement.
He may add to this trust in later years. These assets will be held sepa-
rately, and once the trust is established, they are no longer part of
Bouderi’s portfolio.
■ Give a global balanced mandate to the manager giving its “best effort” to sat-
isfy the risk/return investment goals of the client.
■ Set a specific benchmark for the manager, as done by institutional clients
who use several specialized managers.
■ List the asset classes that can be considered, thereby excluding various types
of investments (e.g., derivatives, emerging markets, shares of “unethical”
firms, etc.).
545
■ Set constraints on the allocation to be followed at all times. For example, the
IPS could state that no more than 40 percent of the portfolio should be
allocated to international investments and that the foreign currency expo-
sure should not exceed 20 percent of the portfolio.
■ Outline the investment philosophy to be followed by the manager.
Some IPSs include the strategic asset allocation to be used in the portfolio, but
we maintain a distinction between the IPS and the determination of the strate-
gic asset allocation. This process is addressed in the section on global asset
allocation.
5. Schedule for review and monitoring: Feedback and control of the portfolio are
essential in reaching investment goals. The IPS can include a schedule for
review of the investment performance and of the IPS itself. Performance eval-
uation is a critical step in the review of the portfolio. The client should be
provided with rates of return for the portfolio on a periodic basis
( performance measurement ), but that is not sufficient. Performance attribution
allows us to understand the sources of performance. So many factors can
affect the performance of a global portfolio that global performance
attribution is crucial in understanding why and how the portfolio performed
as it did. Finally, performance appraisal, typically conducted over a long period
of time, should enable the client to judge whether the manager is doing a
good job.
The IPS also could include monitoring and rebalancing guidelines.
Because of price movements, the composition of the portfolio will pro-
gressively deviate from the strategic asset allocation. Rebalancing the portfo-
lio entails transaction costs, and some guidelines have to be established
regarding rebalancing. More generally, changes in capital market expecta-
tions, or in the client’s investment objectives and constraints, would dictate
revisions in the portfolio. The illustration of a typical IPS continues in
Example 4.
Capital Market Expectations
Capital market expectations are expectations about the future distributions of

returns to asset classes, including expected returns, volatility of returns, and
correlation of returns. Formulating capital market expectations is potentially the
most rewarding part of global asset management, but it is also the most difficult.
The formulation process is usually decomposed into three steps:
■ Defining asset classes
■ Formulating long-term expectations used in strategic asset allocation
■ Formulating shorter-term expectations used in tactical asset allocation
546
EXAMPLE 4 JOHN BOUDERI INVESTMENT POLICY STATEMENT
Brennan prepared an initial investment policy statement for John Bouderi’s

review. After several discussions and further fact-finding by Brennan, the fol-
lowing IPS was approved by Bouderi.
SUMMARY
This investment policy statement outlines the goals and objectives of John
Bouderi (the client) and the long-term investment strategy that is proper for
the achievement of his financial goals. These goals include providing a retire-
ment income and meeting several near-term spending plans.
To fund his retirement, Bouderi sold his share of the family business. The
investment returns from this portfolio are his sole means of support.
In order to achieve a comfortable retirement, this statement will outline
the return-and-risk objectives, acceptable asset classes for inclusion in the port-
folio, portfolio strategies, and any constraints on achieving these goals.
OBJECTIVES
Return Requirement
A return requirement of 9.6 percent is established.
A total return of 4.7 percent after inflation, taxes and investment expenses
is required to meet the anticipated annual spending needs of $90,000 in cur-
rent dollars. Taking into account expected annual inflation of 3 percent and a
tax rate of 15 percent gives a required nominal total return of 9.1 percent on
an annualized basis, before expenses. The return requirement of 9.6 percent
reflects an initial estimate of annual transaction costs and investment manage-
ment expenses of 50 basis points.
Such a return is expected to meet after-tax income needs in real terms over
an extended time horizon. A return of approximately 1 percent in excess of this
amount is desirable to provide modest long-term growth, but is not required.
Risk Tolerance
The risk tolerance of the client is best described as average to slightly below
average. To meet the return requirement, it is understood that risk must be
taken. Although a fluctuation in portfolio value is expected, the portfolio
should be constructed to minimize the likelihood that the portfolio declines by
more than 10 percent in any one year. The achievement of a stable portfolio
with predictable returns is a highly desired outcome.
CONSTRAINTS
Time Horizon
Bouderi has a two-stage time horizon:
■ The next year, defined by several substantial liquidity events, outlined below.
■ The remainder of his retirement period, expected to be at least twenty
years.
547
Liquidity
The following liquidity events are expected to occur in the next year:
■ $50,000 outflow for a European trip
■ $100,000 outflow for a trust for future grandchildren
■ $480,000 net inflow from sale and purchase of homes
No liquidity events past the next year are known to exist, apart from meeting
annual spending requirements.
Tax Issues
The client faces a maximum tax rate of 15 percent on all income, regardless of
the source. He has no outstanding tax liabilities or unresolved issues with the
taxing authorities.
Legal and Regulatory Issues
The client faces no material legal or regulatory constraints; his desire to estab-
lish a trust for his grandchildren is not expected to present any unusual prob-
lems or concerns.
Unique Circumstances
■ The timing of establishing the trust for Bouderi’s grandchildren,
though material, is largely indeterminate.
■ The anticipated exchange of homes, though material, is largely indeter-
minate.
ACCEPTABLE INVESTMENTS
In order to meet the return objective, the following asset classes have been
identified as acceptable within the framework of a diversified portfolio:
■ Australian equities
■ Debt instruments of the Australian government
■ Debt instruments of Australian corporations
■ Equity investments in the United States
■ Equity investments in Europe
■ Money market instruments
Both active and passive strategies may be employed. It is understood that active
strategies may employ higher risk in order to exceed a benchmark return and
that results may not meet expectations.
It is understood that investments outside Australia add currency risk to the
portfolio. Hedging strategies designed to mitigate this risk are acceptable.
548
FREQUENCY OF REVIEW
This investment policy statement and the resulting strategic asset allocation
should be reviewed periodically to ensure that they remain appropriate to the
client’s needs and circumstances. They should be reviewed again if a material
event occurs. Material events would include, but not be limited to, the following:
■ The exchange of homes, resulting in net proceeds substantially differ-
ent than the anticipated $480,000 inflow
■ Bouderi’s health deteriorating markedly
Quantified performance evaluation will be provided detailing the contribution
of currency risk and asset allocation.
The client further understands the importance of his understanding the
investment strategies adopted and the role his ability and willingness to toler-
ate risk will play in meeting his objectives.
Defining Asset Classes

An asset class is a set of “homogeneous” securities, those whose prices are affected by a
common factor. The segmentation of asset classes is usually based on various criteria:
■ Asset type (e.g., debt, equity, real estate)
■ Geography (e.g., domestic vs. international, or regional breakdown)
■ Sector (e.g., technology stocks, energy stocks, intermediate bonds, high yield
bonds)
■ Style (e.g., growth vs. value stocks)
The segmentation and its number of categories depend on the size of the portfolio
and the characteristics of the investor. An institutional investor with a long horizon
can include private equity and other illiquid alternative investments, but this would
not be the case for an elderly individual investor with little wealth. Individual investors
tend to consider few asset classes, and a typical breakdown could include domestic
equity, international equity, domestic bonds, international bonds, domestic real estate,
and cash. A large institutional investor usually considers a more detailed breakdown,
which could include U.S. equity, European equity, Asian equity, emerging-market
equity, investment-grade domestic bonds, high-yield domestic bonds, international
bonds, inflation-linked bonds, emerging-market debt, and so on.
Long-Term Capital Market Expectations: Historical Returns

Long-term capital market expectations, perhaps over the next five or ten years, are
now used to determine the strategic asset allocation of a portfolio. Several approaches
549
are used. As suggested by Maginn et al. (2007),3 two basic approaches are used to
formulate long-term expectations: historical returns and forward-looking returns.
Historical records for mean returns, volatility, and correlation can simply be
projected to repeat in the future; however, there are many limitations in using past
returns to directly formulate expectations on future returns.
Economic conditions in the past, especially the distant past, may not be rele-
vant for the future. A market that has done exceptionally well (or poorly) in the
past because of some specific events (e.g., liberalization of the economy) may not
do so in the future because that specific event will not repeat. The exceptional
equity risk premium in the twentieth century (6 percent for U.S. stocks) was caused
by two factors:
■ Earnings per share grew steadily.
■ Valuation multiples, such as the price/earnings ratio, grew dramatically over
time.
Stock prices went up in part because of real growth, but more importantly, because
valuation multiples rose impressively until 2000. To expect a similar equity risk pre-
mium in the future, an analyst must make the assumption that valuation multiples
will continue upward to unprecedented levels.4
Another problem is the quality of past data. For example, many countries had
fixed-income markets that were controlled by their local governments until the
1980s. This was typically the case for Continental Europe. Bond yields and cash
rates were not market rates but rates that governments regulated so that they could
borrow cheaply. Various capital and currency controls, as well as regulations of
institutional investors, forced local investors to purchase fixed-income securities
with low returns. This is an explanation for the negative historical real yields in
many countries. Fixed-income markets are now open worldwide.
Problems also arise when looking at historical volatility and correlation.
Correlation across equity and bond markets in the past is likely to change over time
because of market integration worldwide. The problem in using unadjusted histor-
ical estimates of volatility and correlation is even more acute for alternative invest-
ments. Some assets trade infrequently. This is the case for many alternative assets
that are not exchange traded, such as real estate or private equity. This is also the
case for illiquid exchange traded securities or over-the-counter instruments often
used by hedge funds. Because prices used are not up-to-date market prices, they
tend to be smoothed over time.5 For example, the appraisal process used in real
estate introduces a smoothing of returns because properties are appraised only
infrequently. The internal rate of return methodology used in private equity also
introduces a smoothing of returns. The infrequent nature of price updates in the
alternative asset world induces a significant downward bias to the measured risk of
the assets. In addition, correlations between alternative investment returns and
3
See “Capital Market Expectations,” in Maginn et al. (2007).
4
An interesting discussion is provided by Arnott and Bernstein (2002).
5
See Terhaar, Staub, and Singer (2003).
550
conventional equity and fixed-income returns, and among the alternative asset
returns themselves, are often close to zero because of the smoothing effect and the
absence of market-observable returns. The bias can be large, so that the true risk is
larger than the reported estimates. Traditional indexes of hedge fund returns also
suffer from a serious survivorship bias (only good-performing hedge funds are
included in the database) that biases historical return measures for this asset class.6
Long-Term Capital Market Expectations: Forward-Looking Returns

Although the past can be useful to forecast the future, simply extrapolating past
risk–return measures is not sufficient. A useful approach is to derive long-term
estimates that would be consistent with market equilibrium. In an integrated world
financial market, expected returns should reflect the risks borne by investors.
Although events could create attractive investment opportunities in the short run,
long-term capital market expectations that are consistent with market equilibrium
are useful to guide investment strategy. An interesting approach has been
developed by Singer and Terhaar (1997) and Terhaar, Staub, and Singer (2003).
This approach uses the international capital asset pricing model (ICAPM) as an
anchor but takes into account market imperfections, such as segmentation and
illiquidity. The approach can be summarized in three steps:
■ Calculate an updated covariance matrix7 (volatility and correlation of asset
classes).
■ Use the ICAPM to infer expected returns for each asset class.
■ Adjust expected returns for possible market segmentation and illiquidity.
The historical covariance matrix serves as the starting point, but it is adjusted to
reflect some of the biases mentioned previously for infrequently traded assets and
is updated to be forward-looking. A multifactor updating approach ensures consis-
tency of the updated matrix.
The expected return on an asset class is equal to the risk-free rate plus a risk
premium. The ICAPM states that the risk premium on any asset class should be
proportional to its beta with the world market portfolio:
RPi = b iMRPM (1)
where
RPi is the risk premium for asset class i
b iM is the beta of asset class i
RPM is the risk premium on the world market portfolio
6
See Asness, Krail, and Liew (2001).
7
Remember that the covariance between two assets is equal to the product of the standard deviation
(volatility) of the two assets times their correlation. So, the covariance matrix contains all risk informa-
tion on asset classes.
551
We assume that the risk premium on any currency is equal to zero. This assumption
could be relaxed if we assume that a specific currency is undervalued and will revert to
its fundamental purchasing power parity (PPP) value in the long run.
Plotting the risk premiums versus the betas would result in all the points lying
on a straight line. Equation 1 can be rewritten by replacing biM by its value:
si
b iM = riM
sM
si RPM
RPi = riM RP = riM si (2)
sM M sM
or
RPi RPM
= ri M (3)
si sM
where si and sM are the standard deviations of returns of asset class i and of the
market portfolio, and riM is the correlation of returns between asset class i and the
market portfolio.
Equation 3 simply states that the expected reward-to-risk ratio (RPi/si), the
Sharpe ratio of an asset class, is equal to the Sharpe ratio of the market portfolio
times the correlation of the asset class with the market portfolio. The lower the cor-
relation, the lower the Sharpe ratio of an asset class. In a fully integrated world, the
risk premium on an asset reflects the risk-diversification property of the asset. An
asset with a correlation of 0.5 with the market portfolio should have a Sharpe ratio
equal to half that of the market. This is because half of the volatility of asset i can be
diversified away and therefore should not be rewarded by a risk premium.
However, world financial markets are not fully integrated, and equilibrium
pricing of some asset classes (e.g., emerging equity markets or private equity)
should reflect their partial segmentation from the world market. To illustrate, let’s
consider an asset class, called emerging market in Example 5, that is fully seg-
mented. The ICAPM does not hold for this asset class. Local investors cannot invest
abroad, and they dominate the local market. They will require a high risk premium
on this asset class because its total risk cannot be diversified away in a global portfo-
lio. The risk premium is set in isolation by local investors, reflecting the total risk si
of the segmented asset class without regard to its diversification ability. If the mar-
ket were integrated in the world market, global investors would set a lower risk pre-
mium (higher price), reflecting the fact that part of the risk of this asset class can
be diversified away in a global portfolio (correlation less than 1). In a way, global
investors can take advantage of segmented markets. In practice, some markets are
partly segmented, so the equilibrium risk premium on these assets should be some-
what higher than what is dictated by a fully integrated ICAPM, but less than under
full segmentation. This is illustrated in Example 5.
Similarly, to hold illiquid assets, investors require compensation in the form of
an illiquidity premium to be added to the risk premium dictated by the ICAPM,
which assumes liquid markets. While the expected return on an illiquid asset can
therefore look high, such an asset has the unpleasant characteristic that any sale of
552
EXAMPLE 5 RISK PREMIUM
Suppose all investors in the world have similar risk aversion and require a
Sharpe ratio of 0.2 on their diversified portfolios. The world market portfolio
has a volatility of 20 percent. An emerging-market asset has a volatility of 30
percent and a correlation of 0.5 with the world market portfolio. The beta of
the emerging-market asset class is
0.5 * 30%>20% = 0.75
1. What is the equilibrium risk premium for the market portfolio?
2. Assuming full integration, what is the equilibrium risk premium for the
emerging market?
3. Assuming full segmentation, what is the equilibrium risk premium for
the emerging market?
SOLUTION
1. The equilibrium Sharpe ratio required for a well-diversified portfolio is
0.2. Hence, the risk premium on the world market portfolio is 4 percent:
RPM RPM
0.2 = =
sM 20%
RPM = 4%
2. If we assume full integration, the ICAPM tells us what the Sharpe ratio
should be for the emerging market:
RPi RPM 4%
= riM = 0.5 * = 0.10
si sM 20%
Hence, RPi = 0.10 * si = 3 percent. This is also equal to beta times the
risk premium on the world market:
0.75 * 4% = 3%
3. If we assume full segmentation, the ICAPM does not hold, and the
emerging market is priced locally, reflecting its total volatility. Local
investors search for a Sharpe ratio of 0.20 for a diversified portfolio, so
RPi
= 0.20
si
Hence, RPi = 0.2 * si = 6 percent.
a sizable order cannot be achieved quickly except at a significant price discount.

Estimating the premium that is required by market participants to compensate for
the illiquidity of an asset class is a difficult task.8
8
Terhaar, Staub, and Singer (2003) calculate illiquidity premiums for a full array of asset classes.
553
These equilibrium expected returns are used to “guide” the formulation of

long-term capital market expectations. Additional forward-looking factors could be
considered to adjust those expectations, but managers should ensure that the long-
term risk and return assumptions are consistent.
Short-Term Capital Market Expectations

The definition of short term varies according to the investor, but long term typically
refers to ten years or more and short term to less than a year or a few years. Changes
in market environment and valuations could suggest that some assets are under- or
overvalued. This can happen when investors have become very optimistic on some
asset class and asset prices have risen well above their perceived intrinsic values.
The opposite can occur when pessimism prevails. Many forecasting and valuation
techniques are used, from subjective to quantitative. Short-term capital market
expectations will suggest (temporary) tactical deviations from the strategic asset
allocation, as discussed in the following section.
Example 6 details the capital market expectations in the John Bouderi case.
EXAMPLE 6 JOHN BOUDERI CURRENCY VIEW AND CAPITAL MARKET

EXPECTATIONS
Brennan’s firm supplies her with some capital market expectations, summarized
below. These were derived from long-term history and a forward-looking model.
Capital Market Expectations
Asset Class Expected Return Standard Deviation
Domestic equity 11% 15%

European equity 14 21
U.S. equity 12 19
Government bonds 6 7
Corporate bonds 7 9
Money market 5 2
A final part of her firm’s analysis confirmed Brennan’s opinion that the
Australian dollar was undervalued and that the undervaluation will get cor-
rected in the long run at a rate of 2 to 3 percent per year. She confirmed her
conclusion by reviewing a report prepared by her firm and based on some IMF
data. The report studied the purchasing power value of the Australian dollar
over the past twenty years. Twenty years ago, the Australian dollar was stable
and regarded as fairly priced relative to the U.S. dollar. In the table that follows,
period 1 refers to twenty years ago and period 2 to the present. The period 1
exchange rate between the U.S. dollar (US$) and the Australian dollar (A$)
554
was A$:US$ = 1.1279, or 1.1279 U.S. dollars per Australian dollar. The period 2
exchange rate had moved to A$:US$ = 0.5106.
Time Period 1 2
End-of-period exchange rate A$:US$ 1.1279 0.5106

CPI Australia 44.4 114.8
CPI U.S. 64.6 116.2
PPP value of A$:US$ 1.1279 0.7847
Assuming that the Australian dollar was fairly priced relative to the U.S.
dollar in period 1, its period 2 value as dictated by PPP can be calculated by
adjusting by the inflation differential between the United States and Australia:
116.2 44.4
PPP value in period 2 = 1.1279 * * = 0.7847
64.6 114.8
The theoretical PPP value of the Australian dollar is US$0.7847, while the
actual spot exchange rate at the end of period 2 was US$0.5106, or a 35 per-
cent undervaluation.
Brennan knows that the possibility of a strong Australian dollar raises the
question of hedging the currency risk of the non-Australian investments. She
wants to hear John’s reaction to the capital market expectations and exchange
rate data she has presented.
JOHN BOUDERI’S REACTION

Bouderi finds the projected returns from outside Australia compelling, but he
is concerned about the increased volatility of European and U.S. markets.
Brennan explains the benefits of adding asset classes with a low correlation to
the Australian market. To support her position, she provides a table of histori-
cal correlation information, along with the expected correlations developed by
the research department of her firm.
Bouderi concurs with Brennan’s rationale for global investing from a diver-
sification standpoint, but he understands that this will add a new risk to his
portfolio due to currency fluctuations. He has read of investors who have lost
substantial sums speculating on currency movements and wonders whether
currency hedging is worth the expense or risk. He expresses doubts about
hedging: “I am familiar with currency fluctuations because Bestbuilt [his com-
pany] buys many of its supplies from Asian firms. Some years it would hurt us,
some years it would help, but in the long run it seemed to balance out. Isn’t the
same true with global investing?”
Brennan replies that fine-tuned risk management is an important compo-
nent of global asset management. If currency fluctuations induce pure risk
without return compensation, it makes sense to hedge to avoid excessive
volatility in portfolio returns. Currency-hedging transaction costs are low.
555
Furthermore, an appreciation of the Australian dollar is likely, and that will

induce a currency loss on foreign investments. Although it is true that currency
fluctuations tend to balance out in the long run, it may take many, many years
before that happens, and Bouderi is starting from a point at which the
Australian dollar seems extremely low. Something like a 50 percent hedging
strategy would be a reasonable middle-of-the-road position, given current mar-
ket expectations and the fact that there is currently no forward premium to
pay. An advantage is that the 50 percent strategy would also minimize regret.
Global Asset Allocation: From Strategic to Tactical
The most important global investment decision is asset allocation. The first step for
a global investor is to decide on a strategic asset allocation (SAA), or the structure of
the portfolio for the long term. This decision is based on long-term capital market
expectations. The process of periodically adjusting asset allocation to reflect
changes in the market environment is referred to as tactical asset allocation (TAA).
Strategic Asset Allocation

The SAA is derived by conducting an asset allocation optimization using long-term
capital market expectations. For portfolio management with individual clients, the
objectives and constraints indicated in the investment policy statement play a big
role in the optimization. In institutional investment management, the SAA often
takes the form of an investable benchmark that is assigned as an objective by the
sponsor to the manager(s). Because the performance of the portfolio is measured
against it, this benchmark provides a strong guide to the manager’s investment
strategy. The question that remains is the choice of the proper global benchmark.
There are three important issues:
■ The scope of the benchmark
■ The set of weights chosen
■ The investor’s attitude toward currency risk
Scope of the Global Benchmark A truly global investor should include all asset
classes, domestic and foreign, in the global benchmark. This global benchmark
could then be broken down into various sub-benchmarks that can be assigned to
different investment managers. In practice, many investors treat domestic and
foreign investments as different asset classes. For example, a Dutch investor could
decide to invest 50 percent of its assets out of the Netherlands, with half invested in
foreign stocks and half in foreign bonds. Then the Dutch investor could assign
some world equity index to the foreign equity manager and some world bond
index to the foreign bond manager. These global indexes should then be
556
calculated by excluding the Netherlands from the indexes. Note that the global
asset allocation will be strongly biased toward Dutch assets because 50 percent of
the assets are invested domestically, so the natural benchmark for the total assets of
the fund will not be a market capitalization–weighted world index. Further note
that within equity, the distinction is often made between investments in developed
and emerging markets, which are usually treated as different asset classes.
Weights in the Global Benchmark A whole range of approaches is used to

determine the benchmark weights. The simplest, most common approach is to use
a published global market index. The weights are proportional to the relative market
capitalizations (caps). A market cap–weighted index is a natural implication of the
theory. It can be easily replicated in a passive strategy because the weights change
in line with price movements in the portfolio. This type of global index is widely
published and is commonly used by institutional investors worldwide. Hence, the
performance can easily be compared across funds and managers.
Some investors are using GDP country weights instead of market-cap country
weights. The idea is interesting because it gives each country a weight proportional
to its economic strength. But implementing this approach is not easy, and it is fairly
costly in a passive portfolio that must be rebalanced each time a new GDP figure is
published or revised in any country. The same problem occurs if a given stock mar-
ket goes up or down while the GDP weights stay constant.
The inclusion of bonds and alternative investments in a variety of currencies
makes the concept of a world market portfolio very difficult to measure and imple-
ment. Investors seldom consider the world market portfolio of all assets a practical
global investment strategy. Most investors exhibit a strong home bias in their invest-
ment strategy. This attitude is often justified, because foreign investments are
regarded as more risky and costly due to lack of information/familiarity, currency
risk, transaction costs, and differential tax treatments. A large investor will therefore
treat each asset class separately, for example, assigning separate benchmarks for
domestic equity, international equity, domestic bonds, and so on. This policy leaves
open the question of the strategic global asset allocation across all asset classes.
Different investor groups should follow different core strategies that reflect their sit-
uations and comparative advantages in terms of costs, taxes, and risks; therefore, pri-
vate and institutional investors may select a customized global strategic asset alloca-
tion. This allocation is then translated into a customized benchmark combining
indexes for each asset class. Studies by Blake, Lehmann, and Timmermann (1999)
and Brinson, Singer, and Beebower (1991) show that the asset allocation decision is
a major determinant of returns on U.K. and U.S. pension plans. So, the weights cho-
sen are of great importance for performance. Optimization techniques are often
used, based on long-term capital market expectations and on characteristics and
constraints of the investor (see the next section).
Institutional investors face constraints in the form of various regulations that
they must follow. But deviating from the peer group also poses a “business” risk. For
example, a pension sponsor that decides to have a much greater international alloca-
tion than its peers is under severe pressure in times when international investments
557
underperform domestic investments. So, the current asset allocation of peers is also a
form of benchmark. The asset allocation of British and Dutch pension funds, for
example, is much more global than that of U.S., German, and Swiss pension funds.
Pension funds in some small countries have a majority of their investments abroad.
The equity allocation also differs markedly, with British funds typically investing some
60 percent of their assets in stocks while French funds typically invest 10 percent. All
of these numbers are for an “average” pension fund; they vary greatly across funds
and change over time. Nevertheless, the obvious picture is that institutional investors
do not follow a uniform investment strategy across the world.
Currency Allocation in the Benchmark Should foreign investments be system-

atically hedged against currency risks? This question has led to an extensive
controversy. The international CAPM (ICAPM) provides useful insights on risk
management. The basic conclusion from theoretical research is that the optimal
portfolio is the world market portfolio partly hedged against currency risk. Partly
hedged means that all foreign assets are optimally hedged when the market is in
equilibrium. We can objectively observe the world market portfolio; we can use
world market caps as weights for the benchmark. Although investment managers
may disagree as to the exact benchmark to use, all of these portfolios approximate
the investable world portfolio and are observable. The problem is identifying the
optimal hedge ratios, because theory tells us that optimal hedge ratios are a
function of the asset to be hedged, and their values depend on unobservable
parameters such as relative preferences of different nationals, risk aversions, and
the net foreign investment position of each country. The existence of currency risk
premiums is central to the determination of optimal hedge ratios. Exchange rates,
like interest rates and stock prices, are financial prices, and risk premiums are
justified. Unfortunately, these currency risk premiums cannot be measured directly
and are likely to be unstable over time. So, pragmatic shortcuts are necessary:
■ A first pragmatic possibility is full hedging. The motivation for a full-hedging
policy is based on the assumption that we cannot tell whether currency risk
premiums are positive or negative; hence, the sole objective is to minimize
the risk. Therefore, we hedge 100 percent and use a fully hedged or unitary-
hedge benchmark as the strategic benchmark.9 Full hedging is simply focus-
ing on minimizing the volatility of the foreign part of the portfolio. This sim-
ple approach has been severely attacked from several angles:
■ First, theory tells us that currency risk premiums should exist if some coun-
tries are net foreign investors (e.g., Japan) or exhibit more risk aversion
than others.
9
Many practitioners use Pérold and Schulman (1988) as a justification for a unitary hedge ratio (100%
hedge). Actually, Pérold and Schulman do consider the case of a zero-currency-risk premium, but they
advocate taking into account the correlation between currency and market risk to determine the “full”
hedge ratio.
558
■ Second, even if we were doing only a passive, risk-minimization hedge and

cared only about risk, not about expected return, we would take into account
the correlation between the currency risk and the asset risk. In the presence
of correlation between currency and market risk, we should use “regression”
hedges, which will generally not equal 1 because of the correlation between
asset returns and currency movements.
■ Third, the relevant measure of risk should not be the volatility of the for-
eign assets taken in isolation, but their contribution to the total risk of the
global portfolio (domestic and foreign). Indeed, currencies provide an ele-
ment of monetary risk diversification for the domestic portfolio. As long as
the proportion of foreign assets in the total portfolio is small (e.g., less
than 10%), the contribution of currency risk is minimal, and it is not worth
the trouble and costs to engage in systematic currency hedging (see
Nesbitt, 1991, and Jorion, 1989).
■ A second pragmatic alternative often proposed is no hedging. This approach

refutes the assumption that full currency hedging reduces the volatility of
return on foreign assets. The question raised is the investor’s time horizon. A
pension fund has a long-term objective and should not be concerned with
short-term risk, such as monthly or quarterly volatility. Given the structure of
its liabilities, a pension fund should instead focus on the risk that a sufficient
return will not be realized over a long horizon, for example, five or ten years.
Froot (1993) shows, theoretically and empirically, that the risk-minimizing
currency hedge is a function of the investment horizon. With a horizon of
ten years, foreign stocks display a greater return volatility when hedged than
unhedged. The reason for this finding is the mean reversion in exchange
rates. Over the short run, currency returns are explained mostly by changes
in the real exchange rate. In the very long run, the purchasing powers of two
currencies tend toward parity, and exchange rate trends are explained
mostly by the inflation differential between the two currencies. Another
motivation for this approach is that systematic currency hedging can be a
costly process. The job of a fund trustee or an investment manager is at stake
in the short run, so whether that person will feel comfortable with measuring
performance and risk solely on such a long horizon is another question.
Also, significant deviations from PPP persist in the long run.
■ A third pragmatic shortcut is to use universal hedge ratios (or arbitrary
hedge ratios). Black (1990) used theory, and some restrictive assump-
tions, to come up with a 0.75 hedge ratio. Gastineau (1995) suggested a
0.5 hedge ratio. Black had to make many extreme assumptions to derive
his universal hedge ratio.10 For example, all countries should have exactly
the same amount of investment abroad (no net foreign investment)
10
For a criticism of this model, see Adler and Solnik (1990) and Adler and Prasad (1992).
559
and no inflation. Also, he postulated that all investors should have an iden-
tical risk aversion, and so on. Even if the principle that each investor should
use exactly the same hedge ratio for every single asset were to be accepted,
the exact value of the universal hedge ratio is still arbitrary because it is based
on a forecast of the future return and volatility of the world market portfolio.
If investors are to make so many “arbitrary” assumptions in order to derive a
universal hedge ratio of, say, 0.75, why not simply assume the result at the
start? In a sense, Gastineau’s “Why bother?” approach is cleaner. He assumes
that 0.5 is the best. Why 0.5? Because it is halfway between 0 and 1, neither of
which is appropriate. A wrong hedging decision can lead to a vast amount of
regret over not having taken the best decision. For example, a U.S. investor
who decided not to hedge currency risk would have incurred a currency loss
of some 40 percent on Eurozone assets from late 1998 to late 2000, with huge
regret at not having been fully hedged. Conversely, a fully hedged U.S.
investor would have missed the 50 percent appreciation of the euro from late
2001 to 2005, again, with huge regret at not having made the “right” hedging
decision. Basically, regret risk stems from a comparison of the ex post return
of the adopted hedging policy relative to the best hedging policy that could
have been chosen. To minimize regret risk, a simple hedging rule would be
to hedge 50 percent. Such a decision will turn out to be almost always wrong
ex post, but the amount of regret will be minimized. Several practitioners
have justified a 50 percent naive hedge ratio on such intuitive grounds.
To summarize, the extent of strategic currency hedging remains an open

theoretical and empirical question. Because hedge ratios differ across assets and
currencies, depending on unobservable foreign asset positions, utility functions,
individual risk aversion, and inflation, no simple practical solution or theoretically
unquestionable benchmark exists, or ever will, for currency allocation. In the
absence of a natural benchmark dictated by theory or systematic empirical observa-
tion, investors have taken various routes. Some use different currency-hedging strate-
gies for different asset classes. For example, the foreign stock benchmark can be
unhedged while the foreign bond benchmark is fully hedged because currency risk is
relatively more important for bonds than for stocks. Currency-hedged global indexes
are now available for both stocks and bonds from the major index providers, so per-
formance of fully hedged portfolios can easily be assessed.
Tactical Asset Allocation

Active managers adjust their asset allocations periodically, typically monthly but
sometimes more often, to reflect changes in the market environment. This strategy is
called dynamic or tactical asset allocation (TAA). TAA is the process of deciding which
asset classes are attractively or unattractively priced, and making short-term departures
from the long-term policy by buying more of the attractive markets and reducing
the holding of unattractive markets. Through this process, the aim is to add value
by achieving a higher return than would be achieved by simply holding the port-
folio defined by the long-term policy. Adjustments made are conditional on new
560
information, so deviations from the SAA are based on short-term developments and
reflect forecasts on market trends in the next few months. This is typically the case
when it is perceived that investors have become too optimistic (pessimistic) on some
asset class, pushing asset prices very high (low) relative to fundamentals. How to
exploit these forecasts depends on the investment philosophy chosen by the
investment manager (see the earlier discussion of various philosophies).
Some managers use a systematic quantitative approach to TAA, in which
adjustments strictly follow some disciplined risk–return optimization process. A set
of variables is used to evaluate the relative attractiveness of markets, namely, their
risk premium. Typical variables used are the level of the dividend yield, the level of
the interest rate, the spread between long-term yields and cash rates, and so on.11
The holdings of an equity market with a low dividend yield (relative to its historical
average) would be adjusted downward on the premise that the market is over-
priced and will soon revert to normality. Rebalancing is done automatically in a
disciplined fashion to prevent emotions or fads from influencing the TAA.
Other managers base their revision on various models of currencies, interest
rates, and equity markets to determine the fundamental value (or fair value) of vari-
ous asset classes (i.e., their fair price). TAA decisions are made if market prices come
to deviate significantly from their fundamental value. While many of the inputs are
quantitative, the TAA decisions are not automatic; they take into account the current
market environment. Efforts are made to understand why those discrepancies have
arisen. Some of these discrepancies could be explained by theories grounded in
behavioral finance. Finding that an asset class is overvalued will not be useful in TAA
unless the market corrects the discrepancy fairly quickly. Other managers simply use
a subjective assessment of changes in the current market environment.
The proposed asset allocation for the John Bouderi case is discussed in
Example 7.
Global Asset Allocation: Structuring and Quantifying

the Process
The global asset allocation process described, with its separation between strategic
and tactical asset allocations, is common among institutional investors such as
pension funds. Private investors make less use of benchmarks and tend to state their
investment objectives in a less formal manner. They sometimes seem to care more
about absolute returns than about deviations from a prespecified benchmark, but
this does not mean that the investment process should not be structured. We will
discuss an adaptable global investment process.
11
There is some evidence that the direction of worldwide stock and bond prices can be predicted to
some extent by using a set of information variables. This fact does not necessarily imply that these
markets are inefficient; predictability could also be explained by a time variation in risk premiums
justified by a change in the socioeconomic environment. In statistical jargon, the strategic asset allo-
cation could be based on long-term, unconditional risk premiums; the tactical asset allocation could
be based on conditional risk premiums.
561
EXAMPLE 7 BOUDERI ASSET ALLOCATION AND HEDGING
Brennan had analyzed possible asset allocations using her firm’s optimization
model. She selected the asset allocation shown below. She chose a mix close to
the efficient frontier and meeting all of Bouderi’s objectives, following the liq-
uidity events.
Proposed Asset Proposed Dollar Expected Standard
Asset Class Allocation Asset Allocation Return Deviation
Domestic equity 30% $570,000 11% 15%

European equity 15 285,000 14 21
U.S. equity 15 285,000 12 19
Government bonds 12 228,000 6 7
Corporate bonds 25 475,000 7 9
Money market 3 57,000 5 2
Total 100% $1,900,000* 9.82% 8%
*Includes proceeds of $480,000 from the home transactions and $150,000 expenses for the trip to Europe and the trust.
Given her capital market expectations, Brennan believes that foreign cur-
rency hedging was in order. She reviews her thinking and explains her reason-
ing to Bouderi:
“Interest rates in Australia, Europe, and the United States are comparable, so
the forward discount/premium is equal to zero. Currency hedging will exactly
offset any loss caused by depreciation of foreign currencies against the Australian
dollar, without having to pay a forward premium. Of course, any potential for-
eign currency gain would also be eliminated. Full hedging would be a natural
strategy, given expectations, but it can take a very long time before currencies
revert to their PPP value. Furthermore, having some foreign currency exposure
in the portfolio could provide some risk-diversification elements in case of a sud-
den surge of inflation in Australia. Also, John, you seem reluctant to do any hedg-
ing. A 50 percent hedge ratio seems a reasonable middle-of-the-road strategy and
could be easily implemented with foreign currency forward contracts.”
Bouderi agrees, and Brennan explains the details. To minimize transaction
costs, the hedge will consist of buying a one-year-maturity forward contract for
Australian dollars equal to half the amount initially invested in U.S. dollars and
euros (selling forward U.S. dollars and euros). The hedge amount will be rebal-
anced yearly. Brennan realizes that this static hedging policy means that the hedge
ratio will diverge from 50 percent as the value of foreign equity moves up or down.
All expected returns provided in Brennan’s mix reflect this hedging strategy.
Apart from the obvious technical and practical problems inherent in investing
abroad, the key issue in global investing is how to structure the asset allocation
process. Essential to this decision process are a variety of uncertain forecasts con-
cerning exchange rates, interest rates, and stock market patterns. The task is
562
further complicated by the fact that many of these variables are, to varying degrees,
interdependent in the global context. For example, all the major stock markets are
linked, but some are more closely linked than others, depending on the integra-
tion of the underlying economies. Global industry factors cut across borders.
Similarly, a change in the interest rate of one currency will affect the exchange
rates and interest rates of other currencies, but not to the same degree. Domestic
asset allocation is simplified by the fact that an investor chooses from a limited vari-
ety of assets, namely, cash, bonds, common stocks, and possibly alternative assets. In
the global context, however, these choices are multiplied by the number of coun-
tries and currencies available, which can, in and of themselves, present certain
practical problems. For example, an investor may be bullish on the Japanese stock
market but not on the yen. The complexity of the global scene, which involves so
many interactions, makes quantification all the more useful and calls for computer
technology. Any added value should be transferred immediately and efficiently
to all accounts, even if they have diverse objectives and constraints as stated in
their IPSs.
This section describes a quantified system or process for portfolio management
based on a top-down approach that is currently used by several global money man-
agement firms, primarily in private banking. Because it is internally consistent, it
avoids the pitfalls sometimes found in the bottom-up approach. The purpose of
this system is to ensure the most efficient use of existing expertise within an organi-
zation and a rapid implementation of new investment ideas in all accounts. The sys-
tem must therefore be structured along the lines of the major common factors
affecting a security’s price behavior. A different model of the world capital market
would lead to a different system. We will discuss here a balanced approach centered
along a country investment philosophy, but an investment philosophy focusing
on worldwide industry factors, or both country and industry factors, could also
be considered.
Our portfolio management system has four major stages: research and market
analysis, asset allocation optimization, portfolio construction, and performance
and risk control. The attraction of the system lies not in its components but in the
way it integrates the four stages, with the aid of computers, to benefit money man-
agers and their clients. The idea is not to generate more reliable forecasts but to
use currently available forecasts better. That goal requires capitalizing most effi-
ciently on the existing expertise within an organization. A diagram of this system is
shown in Exhibit 1.
Research and Market Analysis

To cope with the complexity and rapid changes of the global environment, a
manager must have the technological tools to analyze and interpret large streams
of data. Ideally, everyone involved in the investment decision process—analysts,
investment committee members, and managers—should operate with a common
electronic system (the platform) that allows a free flow of data and research findings
between decision makers.
563
EXHIBIT 1
An Integrated Investment Process
Real -time database updating
Various Macroeconomic Historical

Research qualitative and company risk and return
Department inputs models study
Evaluation of Forecasts and

individual risk estimates for
securities markets and currencies
Investment Qualitative Asset allocation

Policy Committee evaluation optimizer
Optimal asset allocation for

aggressive and defensive
strategies and various types
of accounts
Managers Management assistance
Financial data on manager’s screen

Account structure and computed divergence
from stated objectives and updated asset
allocation policy
Revision suggestions by account and
aggregated by manager
Various models and administrative programs
Global performance analysis

Performance
Evaluation
and
Control Study of performance and risk by account and manager
Study of performance of analysts
Study of performance of the investment policy
Monitoring and periodic adjustments
564
To monitor markets, a large global database with online connections to major

outside databank services is necessary. The database available on the platform
should contain price histories on markets and individual securities, as well as eco-
nomic statistical data. Ideally, the data should cover several previous years and
should be updated daily. All reports and recommendations produced by research
analysts should be available instantly on the platform. Another major use of the
platform is the development and revision of forecasts on currencies, interest rates,
commodity prices, and national stock indexes. The manager should then translate
these forecasts into estimates of total returns in particular base currencies.
Similarly, risk parameters for markets and securities should be available on the
platform and frequently updated.
Asset Allocation Optimization

The objective of any investment strategy is to achieve a superior performance for a
given level of risk. This can be done through subjective discussion among the
members of an investment strategy committee or by using formal optimization
models. Given the investment philosophy selected, an investment strategy should aim
to achieve an efficient asset allocation by regions/country (or currency) and type of
investment (stocks, bonds, money markets). Toward this end, a mean-variance
quadratic program is often used to merge the forecasts and risk estimates generated
by an investment organization. Providing it is optimal, the resulting asset allocation
will outperform a passive strategy only to the extent that the organization’s forecasts
reflect superior expertise in one or more markets, such as stocks, bonds, or currency.
Also, the allocation must conform to the objectives and constraints set out in the IPSs
of the various accounts managed by an organization. Some accounts permit only
equity investment, others impose restrictions on selling short, and still others do
not permit investment in specific asset classes. Plainly, constraints such as these will
affect the potential return on an account, as will the choice of risk level or any
other limitation.
Unless currency risk is systematically hedged, optimal asset allocations will dif-
fer according to the client’s base currency. In theory, managers should care about
real returns, not nominal returns. They should consider the returns calculated in
the currency of the investor and adjusted by the appropriate inflation rate. Because
the volatility of inflation rates is very small compared with that of most asset returns
or currency movements, the use of real or nominal returns would not make much
difference in the results of the optimization procedure.12
The optimization can be conducted on absolute returns. It also can be con-
ducted relative to a prespecified benchmark, with return and risk being measured
in deviation from the benchmark. This is the case when an SAA is first determined
(the benchmark) and tactical revisions are considered based on the current market
12
Some institutional investors take the structure of their liabilities into account to determine the appro-
priate investment strategy. The risk of the assets is measured relative to that of the liabilities rather
than in absolute terms. But the modeling of the liabilities is a difficult task. This process is called
asset/liability management (ALM).
565
environment. The SAA is optimized based on long-term capital market expecta-

tions when the IPS is drafted. The SAA can be periodically revised (e.g., once a
year) following the same process. Tactical deviations from this SAA can be imple-
mented more frequently based on an optimization using short-term capital market
expectations and taking transaction costs into account.
Portfolio Construction
Active security selection in each market is a natural complement to active market
selection. A research department should maintain on the platform an active list of
individual securities in each market. This regularly updated list should provide a
manager with an analyst’s recommendation, possibly in the form of an expected
return in local currency, as well as major risk characteristics of the security, including
sensitivities to various factors and, for bonds, actuarial yields and a measure for
duration. The manager can then use this active securities list to construct the
portfolio according to the asset allocation strategy of a specific client. Multifactor risk
models are necessary to control the active risks taken in security selection.
Managers of a large number of medium-sized accounts find that rebalancing
those accounts to reflect even modest alterations in the client’s investment policy or
the firm’s strategy is extremely time-consuming and unduly repetitive. Reacting to a
major policy or strategy change takes even longer. To cope with this problem, a man-
ager should first value each portfolio in the asset allocation format. Next, each man-
ager should modify the new investment strategy so that it reflects client IPS guidelines
for his major classes of accounts. The extent of this modification will depend on the
type of client, the base currency, and the size of the account. Total transaction costs
and taxes (e.g., on realized capital gains) must be taken into account. Any deviations
of the current asset allocation appear on the manager’s screen. Drawing on the active
securities list, a computer program will make sell-and-buy recommendations for stocks
and bonds that would enable the account to satisfy the new asset allocation guideline.
The manager can either validate the proposed transactions on the screen or make his
own decisions. The market orders and attendant paperwork required for implement-
ing the validated transactions can be generated automatically by the computer. The
use of pooled investment funds simplifies the process.
More quantitatively oriented managers could adjust the sensitivity of the port-
folio to specific market factors in each cell of the asset allocation; for example, a
manager who is bullish on interest rates in Britain could select sterling bonds with a
long duration to increase the sensitivity to a drop in sterling bond yields. Futures
and options can be used to react to sudden threats of a large market movement in
some currency or asset classes. At this level, fine-tuning of the risk management of
the portfolio should come into play.
Performance and Risk Control

The final step in the investment process is to monitor the performance and
risk of individual portfolios. A common problem is that all money managers are
566
outspoken about global risk diversification, but many do not use a structured
allocation process to achieve it. This problem is all the more serious for managers
with active strategies, who tend to concentrate on a few currencies and markets
and are therefore vulnerable to the high risk associated with those currencies
and markets.
Performance control should be driven by an organization’s daily accounting
system. The objective is to be able to answer the following questions about a
portfolio and, in doing so, to assess the effectiveness of its manager:
■ What is the total return on the fund over a specific period?
■ What is the breakdown of the return in terms of capital gains, currency fluc-
tuations, and income?
■ To what extent is the performance explained by asset allocation, market tim-
ing, currency selection, or individual security selection?
■ How does the overall return compare with that of certain benchmarks?
■ Is there evidence of particular expertise in various asset classes and markets?
■ Has the risk-diversification objective been achieved?
■ How aggressive is the manager’s strategy? How does this compare with the
goals of the client?
It should be stressed that the performance of a research department should also be
studied to pinpoint the areas of expertise. This can be done by constructing mock
portfolios based on analyst recommendations for each country and comparing the
subsequent returns with those on the corresponding national indexes. An
illustration of performance evaluation for the John Bouderi case is provided in
Example 8.
In this chapter, we have examined the global investment industry and an inte-
grated investment process for asset management. The performance report and pre-
sentation come at the end of the process, and they provide feedback as the process
is continuously repeated. Because everything in asset management is global, all the
chapters in this book support the process.
EXAMPLE 8 PERFORMANCE EVALUATION OF THE BOUDERI PORTFOLIO
A year has passed, so Brennan meets with John Bouderi for a scheduled review
of the investment policy statement and a review of his portfolio’s performance
for the year. After a general discussion, Brennan moves on to discuss portfolio
performance for the year. She has prepared an informal presentation and sev-
eral performance reports. Based on her experience with Bouderi’s reaction to
quarterly performance reports, she knows that John finds it difficult to under-
stand the sources of international equity performance, although he has
567
become much more willing to accept hedging after all the reading he has done
about it. At their meeting the previous year to set up the strategic asset alloca-
tion, Brennan and Bouderi had agreed on an unhedged benchmark for the
international equity portion of the portfolio. The passive benchmark had a 60
percent weight in the U.S. index and a 40 percent weight in the European
index. Brennan had encouraged Bouderi to follow performance against this
benchmark.
GLOBAL PERFORMANCE EVALUATION

Bouderi’s portfolio has appreciated by 11 percent before taxes and inflation.
Brennan gives him several reports, including the performance report
shown below. Bouderi wants to compare the performance of the unhedged
portfolio with the unhedged benchmark and then reflect on the impact of
currency hedging. The Australian dollar has appreciated over the year
against most currencies by approximately 5 percent. After reviewing the
document, Bouderi feels happy with the performance of his international
portfolio.
PERFORMANCE REPORT FOR INTERNATIONAL PORTFOLIO

Your international equity portfolio rose from $570,000 to $630,285 (a gain of
$60,285 or 10.58%). An additional gain of $14,492 came from the forward cur-
rency contracts in which we sold forward US$79,800 and :79,800 for $285,000
forward. So, the total value of the international portfolio is $644,777, or a total
return of 13.12 percent in Australian dollars. Net of currency hedging, the
return was 10.58 percent, which is only 0.09 percent below the return on your
unhedged benchmark (10.67 percent). All data and calculations are reported
in the tables herein.
The performance of your unhedged international equity portfolio relative to
the benchmark is explained by two factors:
■ Your U.S. equity investments had exactly the same return as the U.S.
index component of the benchmark, namely, 12 percent in Australian
dollars. But your European equity investments overperformed the
European equity index by 0.47% = 9.15% - 8.68%. Because European
equity accounted for 50 percent of your international investments,
the contribution of security selection to total return is a positive
0.24% = 50% * 0.47%.
■ Your asset allocation was more heavily weighted toward European equity
(50% weight) than the benchmark (40% weight). Because European
equity underperformed U.S. equity, this made your international portfolio
underperform the benchmark. The contribution to performance of the
asset allocation decision is -0.33 percent.
568
Hence:13
Portfolio return = Benchmark return + Asset allocation + Security selection
10.58% = 10.67% - 0.33% + 0.24%
The currency hedge that you applied contributed an additional 2.54 percent to
your performance. You sold forward US$79,800 (countervalue $142,500), that
is, half of your U.S. dollar exposure. You did this at a forward rate14 of A$:US$
= 0.56, or US$:A$ = 1.7857 (equal to 1/0.56). The exchange rate has now
moved to A$:US$ = 0.59, or US$:A$ = 1.6949 (equal to 1/0.59). So, you can
buy back the US$79,800 for $135,254, with a gain of $7,246 = 142,500 -
135,254. A similar gain is made on the forward euro sale, and the total gain is
$14,492. Hence, hedging has given an additional return of 2.54% =
14,492/570,000.
The total return on the international equity portfolio is 13.12 percent, well
above your benchmark return of 10.67 percent.
July Current July Current
Index July Previous In Local Currency In A$ Return in A$
United States 100 118 112 12.00%

Europe 100 114.50 108.68 8.68%
Benchmark 100 116.25 110.67 10.67%
A$:US$ 0.56 0.59
A$:: 0.56 0.59
Note: Weights in the benchmark are 60 percent United States and 40 percent Europe.
July Previous July Current Return in Return Contribution

In Local In Local Local in to Portfolio
Currency In A$ Currency In A$ Currency A$ Return
U.S. stocks 159,600 285,000 188,328 319,200 18% 12% 6.00%

Europe stocks 159,600 285,000 183,540 311,085 15% 9.15% 4.58%
Currency 0 14,492 2.54%
futures
Total 570,000 644,777 13.12%
Note: All local currency values are converted at the exchange rates given above. Starting weights in the portfolio are
50 percent U.S. stocks and 50 percent European stocks.
13
Let’s denote R, RUS, and REU as the returns on the portfolio and on the U.S. and European segments;
I, IUS, and IEU as the returns on the benchmark and on the U.S. and European indexes; and SSEU as
the security selection return on European equity. We have
R - I = 0.5RUS + 0.5REU - (0.6IUS + 0.4IEU) = 0.5IUS + 0.5(IEU + SSEU) - (0.6IUS + 0.4IEU)
= (0.5 - 0.6)IUS + (0.5 - 0.4)IEU + 0.5SSEU = -0.33% + 0.24%
14
The forward rate is equal to the spot rate because interest rates are equal in the three countries.
569
Summary
■ There are several types of participants in the global investment arena: investors
(private or institutional), investment managers, brokers, consultants and advis-
ers, and custodians. Some players belong to several categories.
■ Major choices in terms of investment philosophy and strategy must be made
by an investment firm structuring its global investment process. These choices
are based on a view of the global behavior of security prices. A major question
is how active a global strategy should be. In an active strategy, a manager can
decide on (1) global asset allocation by type of asset and currency, (2) security
selection, and (3) market timing. An active manager can also selectively hedge
certain types of risks, such as currency risk. Another question is whether the
major emphasis should be on market analysis (top-down approach) or on
security analysis (bottom-up approach). Yet another important question is
whether the focus should be on regional/ country factors or on global factors
that cut across countries, such as industry factors. These and other choices
dictate how the investment process should be structured.
■ The client’s objectives, constraints, and requirements are specified in an invest-
ment policy statement (IPS), which then forms the basis for the strategic asset
allocation.
■ Capital market expectations are expectations about the future distributions
of returns to asset classes, including expected returns, volatility of returns,
and correlation of returns. Formulating capital market expectations is poten-
tially the most rewarding part of global asset management, but it is also
the most difficult. The formulation process is usually decomposed into
three steps:
■ Defining asset classes
■ Formulating long-term expectations used in strategic asset allocation
■ Formulating shorter-term expectations used in tactical asset allocation
■ The most important global investment decision is the selection of an asset allo-
cation. The first step for a global investor is to decide on a strategic asset alloca-
tion, or the structure of the portfolio for the long term. This choice is based on
long-term capital market expectations. The process of periodically adjusting
asset allocation to reflect changes in the market environment is referred to as
tactical asset allocation.
■ The strategic asset allocation is formalized through one or several benchmarks
set as guidelines to investment managers. Important questions must be
resolved:
■ The scope of the global benchmark
■ The weights in the global benchmark
■ The attitude toward currency risk and hence the currency allocation in the
global benchmark
570
■ Structuring and quantifying the investment process is a difficult task because of

the large number of parameters involved. A disciplined and efficient approach
calls for a partly quantified, integrated system. The objective is to make optimal
use of all expertise and to control risk.
■ For a global investor, multiperiod portfolio performance evaluation is essen-
tial. Careful performance evaluation disentangles attributes and also allows a
comparison of hedged and unhedged positions.
Problems
1. Distinguish between sell-side and buy-side analysts.
2. Describe the difference between a pay-as-you-go retirement system and a capitalized

contribution system.
3. Compare the margins in asset management and investment banking.
4. List four indexing methods, and specify why tracking of a global index is so difficult.
5. Describe four approaches to currency management for portfolios.
6. A retiree, James Timor, has an asset base of :1,900,000. Taking into account annual
inflation of 3 percent, taxes of 15 percent, management fees of 50 basis points, his
objectives, his constraints, and his annual spending needs of :90,000, his financial
advisor recommended an asset allocation with an expected return of 9.82 percent and
a standard deviation of 8 percent. The advisor based this on the following
calculations:
Principal :1,900,000
Return rate 0.0982 186,580
Management expenses 0.005 -9,500
Inflation 0.03 -57,000
Tax rate 0.15 -27,987
Annual spending 92,093
During the next year, Timor’s return is negative 6.18 percent. He has a scheduled meet-
ing with his advisor. If his portfolio’s returns are normally distributed and independent,
what is the probability of a loss of 6.18 percent or more in one year? What is the proba-
bility of three years in a row of returns of less than 9.82 percent?
7. Timor is unnerved a little by losing over :100,000 last year in his portfolio that had
begun the year at :1,900,000. He is now more risk averse and wants to decrease his allo-
cation to stocks and increase his allocation to government securities. What is the effect
of this sequence of events on Timor’s ability to spend :90,000 annually? If his risk toler-
ance had not changed, what advice would have helped him meet his objectives? What
difficulty does his change in risk tolerance create?
8. An individual planning to retire at the end of three years has a defined benefit of
$30,000 per year not protected against inflation. In her planning, she uses a thirty-year
571

GI Book 6e-Trang-2

Uploaded by

Document Informationclick to expand document information

Document Informationclick to expand document information

Copyright:

Available Formats

GI Book 6e-Trang-2

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

GI Book 6e-Trang-2

Uploaded by

Copyright:

Available Formats

Structuring the Global

■ Interpret capital market data and ■ Evaluate the implications of a port-

A n investment organization that is planning to offer global investment products

A Tour of the Global Investment Industry

Institutional Investors The term institutional investor is generally used to

independent boutiques specializing in offering specific investment products.

Consultants and Advisers

companies, endowments, foundations, and high-net-worth individuals. Independent

Global Investment Philosophies

The Passive Approach

weights, which is an impossible task. Indexing requires superior computer, adminis-

The Active Approach

as low-credit-rating bonds and emerging debt. This strategy leads to an

Industry or Country Approach

a manager following the country approach decides on the percentage allo-

EXAMPLE 1 PAIR TRADES

managers resort to different indicators in different countries. Careful analysis is

Still other managers take a proactive tactical approach to currency forecasting.

The Investment Policy Statement

An investment policy statement (IPS) is the cornerstone of the portfolio management

EXAMPLE 2 JOHN BOUDERI CASE CIRCUMSTANCES

RISK AND RETURN OBJECTIVES

IPSs vary in terms of content, especially between private and institutional

assumption of returns being independent and normally distributed, the sam-

EXAMPLE 3 JOHN BOUDERI INVESTMENT CONSTRAINTS

With agreement on investment objectives, Brennan and Bouderi must consider

LEGAL AND REGULATORY ISSUES

Capital Market Expectations

Capital market expectations are expectations about the future distributions of

EXAMPLE 4 JOHN BOUDERI INVESTMENT POLICY STATEMENT

Brennan prepared an initial investment policy statement for John Bouderi’s

Defining Asset Classes

Long-Term Capital Market Expectations: Historical Returns

Long-Term Capital Market Expectations: Forward-Looking Returns

EXAMPLE 5 RISK PREMIUM

a sizable order cannot be achieved quickly except at a significant price discount.

These equilibrium expected returns are used to “guide” the formulation of

Short-Term Capital Market Expectations

EXAMPLE 6 JOHN BOUDERI CURRENCY VIEW AND CAPITAL MARKET

Capital Market Expectations

Asset Class Expected Return Standard Deviation

Domestic equity 11% 15%

End-of-period exchange rate A$:US$ 1.1279 0.5106

JOHN BOUDERI’S REACTION

Furthermore, an appreciation of the Australian dollar is likely, and that will

Global Asset Allocation: From Strategic to Tactical

Strategic Asset Allocation

Weights in the Global Benchmark A whole range of approaches is used to

Currency Allocation in the Benchmark Should foreign investments be system-

■ Second, even if we were doing only a passive, risk-minimization hedge and

■ A second pragmatic alternative often proposed is no hedging. This approach

To summarize, the extent of strategic currency hedging remains an open

Tactical Asset Allocation

Global Asset Allocation: Structuring and Quantifying

EXAMPLE 7 BOUDERI ASSET ALLOCATION AND HEDGING

Domestic equity 30% $570,000 11% 15%

Research and Market Analysis