Final PDF to printer

25
CHAPTER

International

PART VII
Diversification

ALTHOUGH WE IN the United States customarily informational transparency in different countries.

use a broad index of U.S. equities as the market- Therefore, in this chapter we review the major
index portfolio, the practice is increasingly inap- topics covered in the rest of the book, emphasiz-
propriate. U.S. equities represent only about 40% ing their international aspects.
of world equities and a far smaller fraction of total We begin with a brief description of interna-
world wealth. In this chapter, we look beyond tional equity markets, pointing out the wide range
domestic markets to survey issues of interna- of venues available to investors. We then turn to
tional and extended diversification. the central concepts of portfolio theory—risk and
In one sense, international investing may be diversification. We will see how exchange rate
viewed as no more than a straightforward gen- fluctuations add another element to the uncer-
eralization of our earlier treatment of portfolio tainty surrounding rates of return. However, global
selection, with a larger menu of assets from which diversification offers opportunities for improv-
to construct a portfolio. Similar issues of diver- ing portfolio risk–return trade-offs. International
sification, security analysis, security selection, investing also entails a range of political risks. We
and asset allocation face the investor. On the consider several of these risks as well as sources
other hand, international investments pose some of information pertaining to them. Finally, we
problems not encountered in domestic markets. show how active asset allocation can be general-
Among these are the presence of exchange rate ized to incorporate country and currency choices
risk, restrictions on capital flows across national in addition to traditional domestic asset class
boundaries, an added dimension of political risk choices and demonstrate performance attribution
and country-specific regulations, and differing in an international context.

25.1 Global Markets for Equities

You can easily invest today in the capital markets of nearly 100 countries and obtain up-
to-date data about your investments in each of them. By 2015, more than 20 countries had
stock markets with market capitalization above $100 billion.

853

bod77178_ch25_853-[Link] 853 04/18/17 09:00 PM

 Final PDF to printer

854 PART VII Applied Portfolio Management

U.S. investors have several avenues through which they can invest internationally. The
most obvious method, which is available in practice primarily to larger institutional inves-
tors, is to purchase securities directly in the capital markets of other countries. However,
even small investors can take advantage of several investment vehicles with an interna-
tional focus.
Shares of several foreign firms are traded in U.S. markets either directly or in the form
of American depositary receipts, or ADRs. A U.S. financial institution such as a bank
will purchase shares of a foreign firm in that firm’s country and then issue claims to those
shares in the United States. Each ADR is then a claim on a given number of the shares of
stock held by the bank. Some stocks trade in the U.S. both directly and as ADRs.
There is also a wide array of mutual funds with an international focus. In addition to
single-country funds, there are several open-end mutual funds with an international focus.
For example, Fidelity offers funds with investments concentrated overseas, generally in
Europe, in the Pacific Basin, and in developing economies in an emerging opportunities
fund. Vanguard, consistent with its indexing philosophy, offers separate index funds for
Europe, the Pacific Basin, and emerging markets. Finally, as noted in Chapter 4, there
are many exchange-traded funds known as iShares or WEBS (World Equity Benchmark
Shares) that are country-specific index products.
U.S. investors also can trade derivative securities based on prices in foreign security
markets. For example, they can trade options and futures on the Nikkei stock index of 225
stocks traded on the Tokyo stock exchange or on FTSE (Financial Times Share Exchange)
indexes of U.K. and European stocks.
The investments industry commonly distinguishes between “developed” and “emerg-
ing” markets. A typical emerging economy still is undergoing industrialization, is growing
faster than developed economies, and has capital markets that usually entail greater risk.
We use the FTSE1 criteria, which emphasize capital market conditions, to classify markets
as emerging or developed.

Developed Countries
To appreciate the myopia of an exclusive investment focus on U.S. stocks and bonds, con-
sider the data in Table 25.1. The U.S. accounts for only 40.6% of world stock market
capitalization. Clearly, active investors can attain better risk–return trade-offs by extending
their search for attractive securities abroad. Developed countries accounted for about 70%
of world gross domestic product in 2015 and 78.6% of world market capitalization.
Table 25.1 presents data on both market capitalization and GDP for a sample of devel-
oped economies. The United States is by far the largest economy, measured either by GDP
or the size of the stock market. However, Switzerland is the leader in GDP per capita, and
Hong Kong has the largest stock market compared to GDP. On average, the total stock
market capitalization in these countries is about the same magnitude as annual GDP, but
there is tremendous variation in these numbers. Capitalization as a fraction of GDP for the
counties toward the bottom of the list, with smaller capital markets, is generally far lower
than for those counties near the top. This suggests widespread differences in economic
structure even across developed nations.

1
FTSE Index Co. (the sponsor of the British FTSE [Financial Times Share Exchange] stock market index) uses 14
specific criteria to divide countries into “developed” and “emerging” lists.

bod77178_ch25_853-[Link] 854 04/18/17 09:00 PM

 Final PDF to printer

CHAPTER 25 International Diversification 855

Market
Capitalization Percent of GDP per Market Cap
($ billions) World (%) GDP ($ billions) Capita ($) as % of GDP

U.S. $25,068 40.6% $17,947 $55,837 139.7%

Japan 4,895 7.9 4,123 32,477 118.7
Hong Kong 3,185 5.2 310 42,423 1,027.6
U.K. 2,781 4.4 2,849 43,734 97.6
France 2,088 3.4 2,422 36,248 86.2
Germany 1,716 2.8 3,356 41,219 51.1
Canada 1,593 2.6 1,551 43,249 102.8
Switzerland 1,519 2.5 665 80,215 228.6
Korea 1,231 2.0 1,378 27,222 89.4
Australia 1,187 1.9 1,340 56,328 88.6
Spain 787 1.3 1,199 25,832 65.7
Netherlands 728 1.2 753 44,433 96.8
Italy 587 1.0 1,815 29,847 27.5
Belgium 415 0.7 454 40,231 91.3
Mexico 402 0.7 1,144 9,009 35.2
Israel 244 0.4 296 35,330 82.4
Norway 194 0.3 388 74,735 49.9
Chile 190 0.3 240 13,384 79.2
Turkey 189 0.3 718 9,130 26.3
Poland 138 0.2 475 12,494 29.0
Ireland 128 0.2 238 51,290 53.8
Austria 96 0.2 374 43,439 25.7
New Zealand 74 0.1 174 37,808 42.8
Portugal 60 0.1 199 19,223 30.1
Luxembourg 47 0.1 58 101,450 81.6
Greece 42 0.1 195 18,036 21.6
Hungary 18 0.0 121 12,259 14.7
Slovenia 6 0.0 43 20,713 14.1
All developed countries 49,609 78.6 44,823 110.7
World 63,150 100.0 63,982 98.7

Table 25.1
Market capitalization and GDP of developed countries, 2015
Source: The World Bank, [Link].

Emerging Markets
For a passive strategy, one could argue that a portfolio of equities of just the six countries
with the largest capitalization would make up over 70% (in 2015) of the world portfolio
and may be sufficiently diversified. However, this argument will not hold for active port-
folios that seek to tilt investments toward promising assets. Active portfolios will naturally
include many stocks or indexes of emerging markets.
Surely, active portfolio managers do not want to neglect stocks in markets such as China
with an average GDP growth rate in the last five years of 15%. Table 25.2 shows data from

bod77178_ch25_853-[Link] 855 04/18/17 09:00 PM

 Final PDF to printer

856 PART VII Applied Portfolio Management

Market
Capitalization Percent of GDP per Market Cap
($ billions) World (%) GDP ($ billions) Capita (%) as % of GDP

China $8,188 13.3% $10,866 7,925% 75.4%

India 1,516 2.5 2,074 1,582 73.1
South Africa 736 1.2 313 5,692 235.3
Singapore 640 1.0 293 52,889 218.6
Brazil 491 0.8 1,775 8,539 27.6
Russia 393 0.6 1,326 9,057 29.7
Malaysia 383 0.6 296 9,766 129.3
Indonesia 353 0.6 862 3,346 41.0
Thailand 349 0.6 395 5,816 88.2
Philippines 239 0.4 292 2,899 81.8
Colombia 86 0.1 292 6,056 29.4
Peru 57 0.1 192 6,122 29.4
Argentina 56 0.1 583 13,432 9.6
Sri Lanka 21 0.0 82 3,926 25.3
Cyprus 3 0.0 19 22,957 13.9
All emerging markets 13,510 21.9 19,660.6 68.7

Table 25.2
Market capitalization of stock exchanges in emerging markets
Source: The World Bank, [Link].

the largest emerging markets. These markets make up 21% of the world GDP. Per capita
GDP in these emerging markets is quite variable, ranging from $1,582 (India) to $52,889
(Singapore). Market capitalization as a percent of GDP in these countries is still below
70%, suggesting that these markets can grow significantly over the coming years, even
without spectacular growth in GDP.

Market Capitalization and GDP

A contemporary view of economic development2 holds that an important requirement
for economic advancement is a developed code of business laws, institutions, and regula-
tions that allows citizens to legally own, capitalize, and trade capital assets. As a corol-
lary, we expect that development of equity markets will serve as a catalyst for enrichment
of the population.
Figure 25.1 depicts the relationship between per capita GDP and market capitalization
(where both variables have been transformed to log10 scale). The positive slope of the
regression line through the scatter shows that, on average, countries with larger capital
markets also tend to have higher levels of per capita GDP. The slope of the line is .32,
which implies that every percentage point increase in total market capitalization predicts
an increase of .32% in per capita GDP.3 Even so, we observe that most of the observations
2
A highly influential paper in this literature is Rafael La Porta, Florencia Lopez-de-Silanes, Andrei Shleifer, and
Robert W. Vishny, “Law and Finance,” Journal of Political Economy 106 (December 1998), pp. 1113–1155.
3
This simple single-variable regression is not put forward as a causal model but simply as a way to describe the
relation between per capita GDP and the size of markets.

bod77178_ch25_853-[Link] 856 04/18/17 09:00 PM

 Final PDF to printer

CHAPTER 25 International Diversification 857

for the developed countries (the

circles) lie above the regression 100,000
Germany U.S.
line while most of the observa-
Canada
tions for the emerging markets
(the triangles) lie below. This may
Japan
be evidence that legal, regulatory,

Per Capita GDP ($)

10,000
and economic institutions in the Brazil Russia
China
more advanced economies also
contribute to productivity.
India
1,000
Home-Country Bias
We know that purely passive
investment strategies would call
for risky portfolios that are indexed
100
to a broad market portfolio. In the 1 10 100 1000
international context, that mar- Market Capitalization as % of GDP
ket portfolio would include stock
Developed Countries Emerging Markets Regression Line
markets from all nations, and the
market index would be a broadly
diversified world portfolio.
Despite this, it is clear that Figure 25.1 Per capita GDP and market capitalization as percent of
investors everywhere tend to over- GDP, 2015
weight investments in their home Source: The World Bank, [Link].
countries (relative to representa-
tion in the world portfolio) and
underweight investments in foreign assets. For example, referring back to Table 25.1, a
purely indexed U.S. investor would hold about 41% of the total equity portfolio in U.S.
securities with the remaining 59% invested abroad. Yet in practice, we know that U.S. inves-
tors hold a large majority of their equity investments in U.S. firms. This pattern is mirrored
in the behavior of investors around the world and is commonly called home-country bias.

25.2 Exchange Rate Risk and International Diversification

Exchange Rate Risk
When a U.S. investor invests abroad, the dollar-denominated return depends on two factors:
first, the performance of the investment in the local currency, and second, the exchange
rate at which that investment can be brought back into dollars. We illustrate with the fol-
lowing example.

Example 25.1 Exchange Rate Risk

Consider an investment in risk-free British government bills paying 10% annual interest in
British pounds. While these U.K. bills would be the risk-free asset to a British investor, this is
not the case for a U.S. investor. Suppose, for example, the current exchange rate is $2 per
pound and the U.S. investor starts with $20,000. That amount can be exchanged for £10,000
and invested at a riskless 10% rate in the United Kingdom to provide £11,000 in one year.

bod77178_ch25_853-[Link] 857 04/18/17 09:00 PM

 Final PDF to printer

858 PART VII Applied Portfolio Management

What happens if the dollar–pound exchange rate varies over the year? Say that dur-
ing the year, the pound depreciates relative to the dollar, so that by year-end only $1.80 is
required to purchase £1. The £11,000 can be exchanged at the year-end exchange rate for
only $19,800 (= £11,000 × $1.80/£), resulting in a loss of $200 relative to the initial $20,000
investment. Despite the positive 10% pound-denominated return, the dollar-denominated
return is negative 1%.

We can generalize from Example 25.1. The $20,000 is exchanged for $20,000/E0
pounds, where E0 denotes the original exchange rate ($2/£). The U.K. investment grows to
(20,000/E0)[1 + rf (UK)] British pounds, where rf (UK) is the risk-free rate in the United
Kingdom. The pound proceeds ultimately are converted back to dollars at the subsequent
exchange rate E1, for total dollar proceeds of 20,000(E1/E0)[1 + rf (UK)]. Therefore, the
dollar-denominated return on the investment in British bills is

​1 + r (US) = [1 + r​ ​  f​​  (UK)]​E​  1​​/​E​  0​​​ (25.1)

We see in Equation 25.1 that the dollar-denominated return equals the pound-
denominated return times the exchange rate “return.” For a U.S. investor, the invest-
ment in British bills is a combination of a safe investment in the United Kingdom and a
risky investment in the performance of the pound relative to the dollar. Here, the pound
fared poorly, falling from a value of $2 to only $1.80. The loss on the pound more than
offset the earnings on the British bill.
Figure 25.2 illustrates this point. It presents rates of return on stock market indexes in
several countries for 2015. The top bar in each pair depicts returns in local currencies,
while the bottom bar depicts returns in U.S. dollars, adjusted for exchange rate move-
ments. It’s clear that exchange rate fluctuations over this period had large effects on dollar-
denominated returns in several countries.

Concept Check 25.1

Using the data in Example 25.1, calculate the rate of return in dollars to a U.S. investor holding the British bill if
the year-end exchange rate is: (a) E1 = $2.00/£; (b) E1 = $2.20/£.

Exchange rate risk arises from uncertainty in exchange rate fluctuations. The investor
in safe U.K. bills in Example 25.1 still bears the risk of the U.K./U.S. exchange rate. We
can assess the magnitude of exchange rate risk by examining historical rates of change in
various exchange rates and their correlations.
Table 25.3, Panel A, presents the annualized standard deviation of monthly percent
changes in the exchange rates of major currencies against the U.S. dollar over the period
2011–2016. The data show that currency volatility can be quite high. With the excep-
tion of the Chinese RMB, the standard deviation of the percent changes in the exchange
rate was around three-quarters that of the S&P 500, with annualized values ranging from
7.47% (Canadian dollar) to 10.01% (Swiss franc). The annualized standard deviation of
returns on U.S. large stocks for the same period was 12%. An active investor who believes
that a foreign stock is underpriced but has no information about any mispricing of its
currency should consider hedging the currency risk exposure when tilting the portfolio
toward that stock.

bod77178_ch25_853-[Link] 858 04/18/17 09:00 PM

 Final PDF to printer

CHAPTER 25 International Diversification 859

U.S.
Spain
South Korea
Singapore
Russia
Mexico
Japan
Italy
India
Hong Kong
Greece
Germany
France
China
Canada
U.K.
Brazil
-40% -30% -20% -10% 0% 10% 20%

Return (in U.S. $) Return (in Local Currencies)

Figure 25.2 Stock market returns, 2015

Source: The Economist, January 2, 2016.

A. Monthly Change in Exchange Rate against U.S. Dollar

Switzerland China Canada

Euro (€) U.K. (£) (SFr) Japan (¥) (RMB) (C$)

Standard deviation (annualized) 8.19% 8.72% 10.01% 9.47% 2.60% 7.47%

Correlation with S&P 500 −0.095 −0.090 −0.042 0.134 −0.373 −0.453
B. Correlation Matrix of Monthly Changes in Exchange Rate against U.S. Dollar
Switzerland China Canada
Euro (€) U.K. (£) (SFr) Japan (¥) (RMB) (C$)

Euro (€) 1.00

U.K. (£) 0.56 1.00
Switzerland (SFr) 0.43 0.28 1.00
Japan (¥) 0.20 0.23 0.28 1.00
China (RMB) −0.05 0.27 0.09 −0.10 1.00
Canada (C$) 0.46 0.44 0.14 0.14 0.29 1.00

Table 25.3
Exchange rate volatility, 2011–2016
Source: Authors’ calculations using data downloaded from Datastream.

bod77178_ch25_853-[Link] 859 04/18/17 09:00 PM

 Final PDF to printer

860 PART VII Applied Portfolio Management

On the other hand, exchange rate risk may be mostly diversifiable. This is evident both
from the low correlation between exchange rate changes and the return on the S&P 500
shown in the second line of Table 25.3, Panel A, as well as the low correlation coefficients
among currency pairs that appear in Panel B.
Investors can hedge exchange rate risk using a forward or futures contract in foreign
exchange. Recall that such contracts entail delivery or acceptance of one currency for
another at a stipulated exchange rate. To illustrate, recall Example 25.1. In this case, to
hedge her exposure to the British pound, the U.S. investor would agree to deliver pounds
for dollars at a fixed exchange rate, thereby eliminating the risk involved with the eventual
conversion of the pound investment back into dollars.

Example 25.2 Hedging Exchange Rate Risk

If the forward exchange rate in Example 25.1 had been F0 = $1.93/£ when the investment
was made, the U.S. investor could have assured a riskless dollar-denominated return by
arranging to deliver the £11,000 at the forward exchange rate of $1.93/£. In this case, the
riskless U.S. return would have been 6.15%:

​[1 + r​ ​  f​​ (UK)]​F​  0​​  / ​E​  0​​ = (1.10)1.93/2.00 = 1.0615​

The hedge underlying Example 25.2 is the same strategy at the heart of the spot-futures
parity relationship, first discussed in Chapter 22. In both instances, futures or forward
markets are used to eliminate the risk of holding another asset. The U.S. investor can lock
in a riskless dollar-denominated return either by investing in United Kingdom bills and
hedging exchange rate risk or by investing in riskless U.S. assets. Because investments in
two riskless strategies must provide equal returns, we conclude that [1 + rf (UK)]F0/E0 =
1 + rf (US), which can be rearranged to

​F​  ​​ 1 + ​r​  f​​  (US)

​​ ___0 ​ = _________
​   ​​ (25.2)
​E​  0​​ 1 + ​r​  f​​  (UK)

This relationship is called the interest rate parity relationship or covered interest arbi-
trage relationship, which we first encountered in Chapter 23.
Unfortunately, such perfect exchange rate hedging usually is not so easy. In our exam-
ple, we knew exactly how many pounds to sell in the forward or futures market because the
pound-denominated return in the United Kingdom was riskless. If the U.K. investment had
not been in bills, but instead had been in risky U.K. equity, we would have known neither
the ultimate value in pounds of our U.K. investment nor how many pounds to sell forward.
The hedging opportunity offered by foreign exchange forward contracts would thus be
imperfect.
To summarize, the generalization of Equation 25.1 for unhedged investments is that

​1 + r(US) = [1 + r (foreign)]​E​  1​​  / ​E​  0​​​ (25.3)

where r(foreign) is the possibly risky return earned in the currency of the foreign investment
and exchange rates are direct quotes ($ per unit of foreign currency). You can set up a per-
fect hedge only in the special case that r(foreign) is actually known. In that case, you must
sell in the forward or futures market an amount of foreign currency equal to [1 + r (foreign)]
for each unit of that currency you purchase today.

bod77178_ch25_853-[Link] 860 04/18/17 09:00 PM

 Final PDF to printer

CHAPTER 25 International Diversification 861

Concept Check 25.2

How many pounds would the investor in Example 25.2 need to sell forward to hedge
exchange rate risk if: (a) r (UK) = 20%; and (b) r (UK) = 30%?

Investment Risk in International Markets

While active-strategy managers engage in both individual-market asset allocation as well
as security selection, we will restrict our focus to market-index portfolios across countries,
keeping us on the side of an enhanced passive strategy. Nevertheless, our analysis illus-
trates the essential features of extended active management as well.
As we pointed out in Chapter 5, estimates of mean returns are extremely unreliable
without very long data series. Little can be learned about average rates of return from
periods as short as 5 or even 10 years. This is because estimates of expected return essen-
tially depend on only the initial and final stock price (which fully determine the average
rate of increase). Long sample periods are intrinsically necessary for precise estimates of
expected return. However, over such long periods, the mean of the return distribution can
change. This is a difficult hurdle for empirical research.
In contrast, estimates of volatility can be informed by return variation within the sample
period. Increasing the frequency of observations therefore can increase the accuracy of
risk estimates, meaning that precise estimates are feasible even with relatively short sam-
ple periods.
Therefore, we will focus largely on the risk of international investments, where we are
more confident in our empirical estimates. We will use monthly data on country index
returns for the five years ending in October 2016.
Table 25.4 shows key statistics for the investment risk of various equity indexes during
this period. The indexes include most of the world’s largest equity markets as well as mar-
kets from regions with smaller and less developed capital markets. The first pair of col-
umns show the standard deviation of monthly returns measured both in local currency as
well as translated into U.S. dollars, thus reflecting the impact of exchange rate movements
in each month.4
The U.S. equity market (here taken to be the S&P 500) had the lowest volatility in terms
of U.S. dollar-denominated returns over this period, but Canada had the lowest volatility in
local currency. Its higher dollar-denominated volatility reflects the added risk attributable
to exchange rate movements. But in general, the impact of exchange rate movements on
the volatility of dollar-denominated returns is minimal. On average, the volatility of dollar-
denominated returns is not much higher than that of local returns. (The mean and median
estimates reported in Table 25.4 are computed only over the individual country indexes.)
Correlations with the MSCI World index, shown in the next pair of columns, are also
broadly similar for local and dollar-denominated returns. Not surprisingly, given the prom-
inence of the U.S. in the World portfolio and the prominence of its economy in the world
economy, the U.S. has the highest correlation with the World index. Similarly, the FTSE
index (U.K.) and Euronext index (E.U.), both of which measure returns of large and devel-
oped economies, also have very high correlations with the World index, approximately .85
4
Currency-translated returns are not available for the regional indexes. We compute the dollar-denominated return
on the MSCI World portfolio as the weighted average of the dollar-denominated returns of the individual country
indexes in Table 25.4 where weights are based on the proportion of each country in the World portfolio from
Table 25.1. While this sample is not the complete World portfolio, it accounts for over 95% of world market
capitalization.

bod77178_ch25_853-[Link] 861 04/18/17 09:00 PM

 Final PDF to printer

862 PART VII Applied Portfolio Management

Table 25.4
Standard Deviation Correlation with
Stock market of Monthly Returns $-Denominated MSCI World
volatility using both
local and dollar- Local Dollar- Local Dollar-
denominated returns Currency Denominated Currency Denominated

S&P 500 0.032 0.032 0.958 0.905

Nikkei 0.060 0.044 0.696 0.796
FTSE 0.038 0.044 0.863 0.855
Shanghai 0.071 0.072 0.254 0.545
Euronext 0.045 0.048 0.837 0.856
Hang Seng 0.052 0.052 0.653 0.777
Toronto 0.025 0.037 0.711 0.401
Swiss 0.044 0.039 0.726 0.845
India 0.045 0.065 0.614 0.603
Korea 0.035 0.051 0.693 0.743
MSCI-Arabian 0.047 0.494
MSCI-Latin America 0.070 0.725
MSCI World 0.035 0.034 0.952 1.000
Mean 0.045 0.048 0.701 0.733
Median 0.045 0.046 0.703 0.786

Source: Authors’ calculations using returns downloaded from Datastream.

using either dollar-denominated or local currency returns. The Shanghai index is notable
for its low local-currency correlation with the rest of the world.
Table 25.5 shows results for index model regressions of each country’s index against
the MSCI World portfolio. The betas of each index against the World portfolio appear
in the first pair of columns. The beta of the U.S. against the World portfolio is less than
1, as is the median beta of the other indexes in the sample. Betas computed from dollar-
denominated returns are on average higher than those using local returns but, with the
exception of China, not substantially so.
We do not report country-index alphas in this table because results over 5 years are such
unreliable forecasts of future performance. As we noted above, this is a common problem
in using an historical sample to estimate expected returns (as opposed to risk measures).
It is why many decades of data are commonly used to estimate “normal” returns on the
broad market portfolio. Nevertheless, in this sample period, we can report that the median
alpha in local currency was only 1 basis point, almost precisely zero. The dollar generally
appreciated over this period, so the average dollar-denominated alpha was negative, −32
basis points per month.
The second pair of columns in Table 25.5 show the residual standard deviation for each
country index. As discussed in Chapter 8, the regression residual in each month is the
portion of the country return that is independent of the return on the World portfolio. The
entries in these columns are therefore estimates of the standard deviation of “country-
specific returns.” The results are consistent with the other risk measures. The U.S. has
by far the lowest country-specific risk, reflecting its prominence in the World portfolio,
and China has the highest, consistent with its lower correlation with the World portfolio.
Indexes for the countries in the Arab world and Latin America have above-average nonsys-
tematic risk despite the fact that these indexes already enjoy some diversification across

bod77178_ch25_853-[Link] 862 04/18/17 09:00 PM

 Final PDF to printer

CHAPTER 25 International Diversification 863

Table 25.5
Beta against Residual Standard
MSCI World Deviation Index model regres-
sions of country
Local Dollar- Local Dollar- indexes against the
Currency Denominated Currency Denominated MSCI World index
using both local and
S&P 500 0.885 0.847 0.009 0.014
dollar-denominated
Nikkei 1.207 1.022 0.043 0.027 returns
FTSE 0.955 1.095 0.019 0.023
Shanghai 0.521 1.150 0.069 0.061
Euronext 1.089 1.195 0.025 0.025
Hang Seng 0.975 1.186 0.039 0.033
Toronto 0.512 0.428 0.018 0.033
Swiss 0.919 0.954 0.030 0.021
India 0.804 1.140 0.036 0.052
Korea 0.692 1.109 0.025 0.034
MSCI-Arabian 0.667 0.041
MSCI-Latin America 1.473 0.049
World 1.000 0.000
Mean 0.856 1.013 0.031 0.032
Median 0.902 1.102 0.028 0.030

Source: Authors’ calculations using returns downloaded from Datastream.

the countries included in each region. Average 100

country-specific risk is essentially the same
regardless of whether returns are measured
80
in local or foreign currency. By and large,
country-specific risk is meaningful. Even the
U.S. value of 1.4% per month is not trivial. The
Percent Risk

60
message is that international diversification is
potentially valuable.
By and large, the results in Table 25.4 and 40
Table 25.5 indicate that investment risk is U.S. Stocks
27
pretty much the same regardless of whether
20
we use local currency or dollar-denominated Global Stocks
11.7
returns. Therefore, we will focus largely on
local returns when we turn to the potential for 0
international diversification in the next section. 1 10 20 30 40 50
Number of Stocks
International Diversification
In Chapter 7, we looked at the risk of equally Figure 25.3 International diversification. Portfolio standard
weighted portfolios composed of different deviation as a percentage of the average standard deviation
numbers of U.S. stocks chosen at random of a one-stock portfolio
and saw the efficacy of naїve diversification. Source: B. Solnik, “Why Not Diversify Internationally Rather Than Domestically?”
Figure 25.3 presents the results of a simi- Financial Analysts Journal, July/August 1974, pp. 48–54. Copyright 1995,
CFA Institute. Reproduced and republished from Financial Analysts Journal
lar exercise, but one in which diversification with permission from the CFA Institute.
includes stocks from around the world. You can

bod77178_ch25_853-[Link] 863 04/18/17 09:00 PM

 Final PDF to printer

Investors’ Challenge: Markets Seem Too Linked

WORDS FROM THE STREET

It’s one of the golden rules of investing: Reduce risk by diver- thought if the investment decisions were made some time
sifying your money into a variety of holdings—stock funds, ago, says Mr. Ezrati, chief economist at money-management
bonds, commodities—that don’t move in lockstep with one firm Lord Abbett & Co. He adds that over long periods of time,
another. And it’s a rule that’s getting tougher to obey. going back decades, sometimes varied asset classes tend
According to recent research, an array of investments to converge.
whose prices used to rise and fall independently are now One explanation for today’s higher correlation is increased
increasingly correlated. For an example, look no further than globalization, which has made the economies of various coun-
the roller coaster in emerging-markets stocks of recent weeks. tries more interdependent. International stocks, even with their
The MSCI EAFE index, which measures emerging markets, higher correlations at present, deserve some allocation in a
now shows .96 correlation to the S&P, up from just .32 six long-term investor’s holdings, says Jeff Tjornehoj, an analyst at
years ago. data firm Lipper Inc. Mr. Tjornehoj is among those who believe
For investors, that poses a troubling issue: how to main- these correlations are a temporary phenomenon, and expects
tain a portfolio diversified enough so all the pieces don’t tank that the diversity will return some time down the line—a year
at once. or few years.
The current correlation trend doesn’t mean investors
should go out and ditch their existing investments. It’s just Source: Shefali Anand, “Investors Challenge: Markets Seem Too
that they may not be “getting the same diversification” they Linked,” The Wall Street Journal, June 2, 2006, p. C1.

see that extending the universe of investable assets to foreign stocks allows even greater
opportunities for risk reduction.
Of course, as emphasized in the nearby box, the benefits from diversification depend
on the correlation structure among securities. The box notes that international correlations
have increased over time. Table 25.6 shows correlation pairs (using local currency returns)
for our sample of country and regional indexes from a recent five-year period. This cor-
relation matrix can be used to construct the minimum-variance portfolio, which provides a

S&P Hang Latin

500 Nikkei FTSE Shanghai Euronext Seng Toronto Swiss India Korea Arabian Am. World

S&P 500 1.000 0.684 0.832 0.231 0.790 0.555 0.633 0.697 0.562 0.621 0.461 0.677 0.958
Nikkei 0.684 1.000 0.644 0.341 0.621 0.484 0.377 0.636 0.511 0.513 0.283 0.473 0.696
FTSE 0.832 0.644 1.000 0.233 0.844 0.651 0.649 0.724 0.588 0.627 0.353 0.716 0.863
Shanghai 0.231 0.341 0.233 1.000 0.232 0.590 0.231 0.113 0.153 0.360 0.199 0.277 0.254
Euronext 0.790 0.621 0.844 0.232 1.000 0.535 0.559 0.730 0.455 0.578 0.339 0.558 0.837
Hang Seng 0.555 0.484 0.651 0.590 0.535 1.000 0.615 0.393 0.548 0.711 0.486 0.753 0.653
Toronto 0.633 0.377 0.649 0.231 0.559 0.615 1.000 0.484 0.521 0.543 0.442 0.674 0.711
Swiss 0.697 0.636 0.724 0.113 0.730 0.393 0.484 1.000 0.339 0.485 0.272 0.355 0.726
India 0.562 0.511 0.588 0.153 0.455 0.548 0.521 0.339 1.000 0.566 0.266 0.624 0.614
Korea 0.621 0.513 0.627 0.360 0.578 0.711 0.543 0.485 0.566 1.000 0.478 0.692 0.693
Arabian 0.461 0.283 0.353 0.199 0.339 0.486 0.442 0.272 0.266 0.478 1.000 0.472 0.494
Latin Am. 0.677 0.473 0.716 0.277 0.558 0.753 0.674 0.355 0.624 0.692 0.472 1.000 0.725
World 0.958 0.696 0.863 0.254 0.837 0.653 0.711 0.726 0.614 0.693 0.494 0.725 1.000

Table 25.6
Correlation matrix of returns using local currency returns, 2011–2016
Source: Authors’ calculations using returns downloaded from Datastream.

864

bod77178_ch25_853-[Link] 864 04/18/17 09:00 PM

 Final PDF to printer

CHAPTER 25 International Diversification 865

Table 25.7
Minimum
Equally Minimum Variance Composition and
Weighted Variance Portfolio volatility of
Portfolio Portfolio (no short sales) internationally
diversified portfolios
A. Weights
S&P 500 0.083 0.318 0.114
Nikkei 0.083 −0.034 0.000
FTSE 0.083 0.207 0.000
Shanghai 0.083 0.051 0.024
Euronext 0.083 −0.149 0.000
Hang Seng 0.083 −0.103 0.000
Toronto 0.083 0.696 0.742
Swiss 0.083 −0.107 0.000
India 0.083 0.046 0.000
Korea 0.083 0.308 0.109
Arabian 0.083 0.044 0.012
Latin Am. 0.083 −0.276 0.000
B. Volatility
Std dev 0.033 0.019 0.024

Source: Authors’ calculations using returns downloaded from Datastream.

better estimate of the potential benefits from diversification than the naїve diversification
exercise conducted in Figure 25.3.
Table 25.7 shows portfolio volatility using equal weights for each country or regional
index (column 1), using minimum variance weights (column 2), or using minimum vari-
ance weights without allowing short sales (column 3). The equally weighted portfolio is
in the spirit of Figure 25.3. Even naïve diversification provides considerable benefit: The
standard deviation of the portfolio is 3.3%, which is only 70% of the average of the indi-
vidual country standard deviations. But we can reduce volatility considerably from this
level. The minimum variance portfolio has a standard deviation of only 1.9% when we
allow short sales, and 2.4% when we do not. We see here ample evidence for the potential
of international diversification to substantially reduce portfolio risk.
What about efficient diversification that achieves the best risk–return trade-off?
This is harder to assess because, as we have pointed out several times, sample average
returns do not provide reliable estimates of expected returns. However, as an alterna-
tive to historical returns, we can estimate expected returns from an international ver-
sion of the CAPM. Table 25.8 presents such estimates using the MSCI World index as
the market portfolio and assuming a risk-free rate of 2% and market risk premium of
8%. Given these expected returns, the standard deviations in Table 25.4, and the cor-
relation matrix in Table 25.6, we can generate the efficient frontier. The results appear
in Figure 25.4.
The frontier in Figure 25.4 is drawn without allowing for short sales. This reflects the con-
straints on shorting imposed on many institutional traders. Even with this restriction, how-
ever, the benefits of international diversification are evident. The U.S. portfolio (the S&P
500) appears as the solid dot and is located substantially below the capital allocation line,

bod77178_ch25_853-[Link] 865 04/18/17 09:00 PM

 Final PDF to printer

866 PART VII Applied Portfolio Management

Table 25.8
Beta Expected Return (%)
Expected rates of return
using local-currency betas S&P 500 0.885 9.1%
against the MSCI World Nikkei 1.207 11.7
portfolio and an interna- FTSE 0.955 9.6
tional CAPM Shanghai 0.521 6.2
Euronext 1.089 10.7
Hang Seng 0.975 9.8
Toronto 0.512 6.1
Swiss 0.919 9.3
India 0.804 8.4
Korea 0.692 7.5
MSCI—Arabian 0.667 7.3
MSCI—Latin America 1.473 13.8
World 1.000 10.0

Source: Betas calculated by authors using returns downloaded from

Datastream. Expected returns are derived from the CAPM, assuming a risk-
free rate of 2% and market risk premium of 8%.

1.4%

1.2%
Risk Premium (% per month)

1.0%
Capital Allocation Line
0.8% Efficient Frontier

0.6% U.S.

0.4%

0.2%

0.0%
2.0% 2.5% 3.0% 3.5% 4.0% 4.5% 5.0%
Standard Deviation of Monthly Return (%)

Figure 25.4 Efficient frontier and CAL using country and regional stock indexes
Source: Authors’ calculations using data from Tables 25.7 and 25.8.

despite the fact that its weight in the tangency portfolio is nearly 50%. The CAL supported
by the U.S. index has a Sharpe ratio (based on monthly returns) of .185, considerably lower
than the Sharpe ratio of the tangency portfolio, which is .248. Thus, a U.S. investor who
foregoes international equity markets is giving up ample opportunities to improve the risk–
return trade-off of his risky portfolio.

bod77178_ch25_853-[Link] 866 04/18/17 09:00 PM

 Final PDF to printer

CHAPTER 25 International Diversification 867

Are Benefits from International Diversification

Preserved in Bear Markets?
Some studies suggest that correlation in country portfolio returns increases during periods
of turbulence in capital markets.5 If so, benefits from diversification would be lost exactly
when they are needed the most. For example, a study by Roll of the crash of October 1987
shows that all 23 country indexes studied declined over the crash period of October 12–26.6
This correlation is reflected in the movements of regional indexes depicted in Figure 25.5.
Roll found that the beta of a country index on the world index (estimated prior to the crash)
was the best predictor of that index’s response to the October crash of the U.S. stock mar-
ket. This suggests a common factor underlying the movement of stocks around the world.
This model predicts that a macroeconomic shock would affect all countries and that while
diversification can mitigate risk, it cannot eliminate exposure to such broad-based events.
The 2008 crash of stock markets around the world allows us to test Roll’s prediction.
The data in Figure 25.6 include average monthly rates of return for both the 10-year period
1999–2008 and the crisis period corresponding to the last 4 months of 2008, as well as the
beta on the U.S. market and monthly standard deviation for several portfolios. The graph
shows that both beta against the U.S. and the country-index standard deviation help explain
the difference between crisis period returns and overall period averages. Market behavior
during the 1987 crisis, that is, larger correlations in extreme bad times, repeated itself in
the crisis of 2008, vindicating Roll’s prediction.

1.05

0.95
Value of One Currency Unit

0.9

0.85
Symbols Positioned
at Market Close
0.8 Local Time
North America
0.75 Ireland, So. Africa, U.K.
Large Europe
Small Europe
0.7
Asia
Australia/New Zealand
0.65
12 14 16 18 20 22 24 26
Tick Marks on October Date, 4:00 P.M., U.S. Eastern Standard Time

Figure 25.5 Regional indexes around the crash, October 14–October 26, 1987
Source: From Richard Roll, “The International Crash of October 1987,” Financial Analysts Journal,
September–October 1988. Copyright 1995, CFA Institute. Reproduced from Financial Analysts Journal with
permission from the CFA Institute.

5
F. Longin and B. Solnik, “Is the Correlation in International Equity Returns Constant: 1960–1990?” Journal of
International Money and Finance 14 (1995), pp. 3–26; and Eric Jacquier and Alan Marcus, “Asset Allocation
Models and Market Volatility,” Financial Analysts Journal 57 (March/April 2001), pp. 16–30.
6
Richard Roll, “The International Crash of October 1987,” Financial Analysts Journal, September–October 1988.

bod77178_ch25_853-[Link] 867 04/18/17 09:00 PM

 Final PDF to printer

868 PART VII Applied Portfolio Management

SD of Monthly Returns, 1999 – 2008 10.0 1.6

9.0 1.4
8.0
1.2
7.0

Beta against U.S.

6.0 1

5.0 SD 0.8
4.0 Beta on U.S.
0.6
3.0
0.4
2.0
1.0 0.2

0.0 0
-17.0 -15.0 -13.0 -11.0 -9.0 -7.0
Deviation from Average Return during Financial Crisis

Average Monthly Return Deviation Beta on

Market 1999–2008 2008: Sept.–Dec. from Average U.S. SD

USA -0.47 -8.31 -7.84 1 4.81

World largest six (non–U.S.) markets -0.16 -7.51 -7.35 0.77 4.71
EU developed markets -0.05 -10.34 -10.29 1.06 6.08
Other Europe developed markets -0.14 -7.59 -7.73 0.82 4.95
Australia + Far East -0.10 -9.29 -9.38 1.04 6.21
Emerging Far East + South Africa -0.20 -9.70 -9.90 1.01 7.10
Emerging Latin America -0.80 -11.72 -12.52 1.27 7.83
Emerging markets in Europe -0.90 -15.43 -16.32 1.38 9.54
World minus U.S. (48 countries by cap) -0.01 -8.79 -8.81 0.91 5.19
World portfolio (by country cap) -0.15 -8.60 -8.45 0.94 4.88

Figure 25.6 Beta and SD of portfolios against deviation of monthly return over September–December 2008
from average return over 1999–2008
Source: Authors’ calculations.

25.3 Political Risk

In principle, security analysis at the macroeconomic, industry, and firm-specific level is
similar in all countries. Such analysis aims to provide estimates of expected returns and
risks of individual assets and portfolios. However, information about assets in a foreign
country is by nature more difficult to acquire.
Consider two investors: an American wishing to invest in Indonesian stocks and an
Indonesian wishing to invest in U.S. stocks. While each would have to conduct funda-
mental analysis of each foreign company, the task would be much more difficult for the
American investor. The reason is not that investment in Indonesia is necessarily riskier
than investment in the U.S. You can easily find U.S. stocks that are, in the final analysis,
riskier than a number of Indonesian stocks. The difference lies in the fact that U.S. finan-
cial markets are more transparent than those of Indonesia.

bod77178_ch25_853-[Link] 868 04/18/17 09:00 PM

 Final PDF to printer

CHAPTER 25 International Diversification 869

In addition, political risk varies dramatically across countries and its assessment
requires expertise in each nation’s economic, legal, tax, and political environment. A lead-
ing organization providing political risk assessment for investors is the PRS Group (Politi-
cal Risk Services), and the presentation here follows the PRS methodology.7
PRS’s country risk analysis results in a country composite risk rating on a scale of 0
(most risky) to 100 (least risky). To illustrate, Table 25.9 shows the rank of a small sample
of countries from the January 2015 issue of PRS’s International Country Risk Guide. It is
not surprising to find Switzerland at the top of the low-risk list and small emerging markets
at the bottom, with Somalia (ranked 140) closing the list. What may be surprising is the
fairly mediocre ranking of the U.S. (ranked 27).
The composite risk rating is a weighted average of three measures: political risk, finan-
cial risk, and economic risk. Political risk is measured on a scale of 100–0, while financial
risk and economic risk are each measured on a scale of 50–0. The three measures are
added and divided by 2 to obtain the composite rating. The variables used by PRS to deter-
mine the composite risk rating from the three measures are shown in Table 25.10.
Table 25.11 shows the three risk measures for 11 countries in order of the January 2015
ranking of composite risk. The table shows that the United States was among the best per-
formers in terms of political risk. But it was at the bottom of this (admittedly short) list in
terms of financial risk. The surprisingly poor performance of the U.S. in this dimension
was probably due to its exceedingly large government and balance-of-trade deficits, which
put pressure on its exchange rate. Exchange rate stability, foreign trade imbalance, and
foreign indebtedness all enter PRS’s computation of financial risk.

Table 25.9
Risk Rating, Risk Rating,
Rank Country 2015 2014 Composite risk ratings
for January 2014 and
1 Switzerland 89.8 89.5 January 2015
6 Canada 83.0 82.3
6 Germany 83.0 85.3
10 Qatar 82.3 82.0
19 Japan 79.5 81.5
27 United States 77.3 75.3
47 China, Peoples’ Rep. 72.0 73.3
59 Spain 69.5 69.3
68 Brazil 68.0 69.3
77 Indonesia 67.0 65.5
90 Russia 64.5 69.5
97 Turkey 63.3 59.3
118 Haiti 60.8 58.8
126 Pakistan 58.3 61.3
132 Venezuela 54.8 54.3
138 Liberia 49.3 52.8
139 Syria 41.3 42.0
140 Somalia 37.5 37.5
Source: International Country Risk Guide, January 2015, The PRS Group, Inc.

7
You can find more information on the Web site: [Link]. We are grateful to the PRS Group for
supplying data and guidance.

bod77178_ch25_853-[Link] 869 04/18/17 09:00 PM

 Final PDF to printer

870 PART VII Applied Portfolio Management

Political Risk Variables Financial Risk Variables Economic Risk Variables

Government stability Foreign debt (% of GDP) GDP per capita

Socioeconomic conditions Foreign debt service (% of exports) Real annual GDP growth
Investment profile Current account (% of exports) Annual inflation rate
Internal conflicts Net liquidity (months of import cover) Budget balance (% of GDP)
External conflicts Exchange rate stability Current account balance (% GDP)
Corruption
Military in politics
Religious tensions
Law and order
Ethnic tensions
Democratic accountability
Bureaucracy quality

Table 25.10
Variables used in PRS’s political risk score

Table 25.11
Country Political Financial Economic Composite
Country risk rankings
by category, 2015 Pakistan 48.5 37.5 30.5 58.3
Turkey 53.0 37.5 36.0 63.3
Russia 56.5 35.5 37.0 64.5
India 61.0 43.0 33.5 68.8
China, Peoples’ Rep. 56.5 47.5 40.0 72.0
United States 82.5 32.5 39.5 77.3
United Kingdom 84.0 35.5 38.0 78.8
Qatar 73.5 42.5 48.5 82.3
Canada 85.0 39.0 42.0 83.0
Germany 84.5 38.0 43.5 83.0
Switzerland 88.0 46.5 45.0 89.8

Source: International Country Risk Guide, January 2015, The PRS Group, Inc.

Finally, Table 25.12 shows ratings of political risk by each of its 12 components. The
U.S. does well in corruption risk (variable F) and democratic accountability (variable K).
China does well in government stability (variable A) but poorly in democratic account-
ability (variable K).
Each monthly issue of the International Country Risk Guide of the PRS Group includes
great detail and holds some 250 pages. Other organizations compete in supplying such
evaluations. The result is that today’s investor can become well equipped to properly assess
the risk involved in international investing.

bod77178_ch25_853-[Link] 870 04/18/17 09:00 PM

 Final PDF to printer

CHAPTER 25 International Diversification 871

This table lists the total points for each of the following political risk components out of the maximum points indicated. The
final column shows the overall political risk rating (the sum of the points awarded to each component).

A Government stability 12 G Military in politics 6

B Socioeconomic conditions 12 H Religious tensions 6
C Investment profile 12 I Law and order 6
D Internal conflict 12 J Ethnic tensions 6
E External conflict 12 K Democratic accountability 6
F Corruption 6 L Bureaucracy quality 4

Risk
Country A B C D E F G H I J K L Rating

Pakistan 6.0 5.5 7.0 6.0 9.0 2.0 1.5 1.0 3.0 1.0 4.5 2.0 48.5
Turkey 7.5 6.0 6.0 7.0 7.0 2.5 2.0 4.0 3.0 2.0 4.0 2.0 53.0
Russia 9.0 5.5 7.5 7.5 7.0 1.5 4.0 5.5 3.0 3.0 2.0 1.0 56.5
China, Peoples’ Rep. 9.0 7.0 6.0 7.0 8.0 2.0 3.0 4.0 3.5 3.5 1.5 2.0 56.5
India 8.5 5.0 7.5 6.5 9.0 2.5 4.0 2.5 4.0 2.5 6.0 3.0 61.0
Qatar 10.5 8.0 10.0 9.5 8.5 4.0 4.0 4.0 5.0 6.0 2.0 2.0 73.5
United States 7.5 9.5 12.0 10.0 9.5 4.5 4.0 5.5 5.0 5.0 6.0 4.0 82.5
United Kingdom 7.5 9.5 11.5 10.0 9.5 5.0 6.0 6.0 5.0 4.0 6.0 4.0 84.0
Germany 8.5 9.0 11.0 10.5 10.5 5.0 6.0 5.0 5.0 4.0 6.0 4.0 84.5
Canada 7.5 8.5 12.0 10.0 11.0 5.0 6.0 6.0 5.5 3.5 6.0 4.0 85.0
Switzerland 9.0 10.5 11.5 12.0 10.5 5.0 6.0 4.5 5.0 4.0 6.0 4.0 88.0

Table 25.12
Political risk points by component, January 2015
Source: International Country Risk Guide, January 2015, The PRS Group, Inc.

25.4 International Investing and Performance Attribution

Because many security markets may be less efficient than those in highly developed
economies such as the U.S., international investing offers greater opportunities for active
managers. International investing calls for specialization in additional fields of analysis:
currency, country and worldwide industry, as well as a greater universe for stock selection.

Constructing a Benchmark Portfolio of Foreign Assets

Active (as well as passive) international investing requires a benchmark portfolio (called
the bogey). One widely used international market index is the MSCI World index. MSCI
also computes variants on this index that exclude some large markets such as the U.S.
Another popular non-U.S. index is the Europe, Australasia, Far East (EAFE) index of
large and mid-cap securities in 21 developed markets, also computed by MSCI. Other
indexes of world equity performance are published by Capital International Indices,
Salomon Brothers, Credit Suisse First Boston, and Goldman Sachs. Portfolios designed
to mirror or even replicate the country, currency, and company representation of these
indexes would be the obvious generalization of the purely domestic passive equity strategy.

bod77178_ch25_853-[Link] 871 04/18/17 09:00 PM

 Final PDF to printer

872 PART VII Applied Portfolio Management

An issue that sometimes arises in the international context is the appropriateness of

market-capitalization weighting schemes in the construction of international indexes. Cap-
italization weighting is far and away the most common approach. However, some argue
that it might not be the best weighting scheme in an international context. This is in part
because different countries have differing proportions of their corporate sector organized
as publicly traded firms.
Table 25.13 shows market-capitalization weights versus GDP weights for countries in
the EAFE index. These data reveal substantial disparities between the relative sizes of mar-
ket capitalization and GDP. The countries with smaller stock markets (toward the bottom
of the table) tend to have a much smaller share of EAFE total market capitalization than
they do of EAFE GDP. For example, while Portugal accounts for 0.5% of total EAFE GDP,
it represents only 0.1% of the market cap of the index. At the other extreme, Hong Kong
accounts for only 0.7% of GDP, but it represents 6.5% of total EAFE market cap. These
disparities indicate that a greater proportion of economic activity is conducted by publicly
traded firms in Hong Kong than in the other EAFE countries.
Some argue that it would be more appropriate to weight international indexes by GDP
rather than market capitalization. The justification for this view is that an internationally
diversified portfolio should purchase shares in proportion to the broad asset base of each
country, and GDP might be a better measure of the weight of a country in the international

Table 25.13
% of EAFE Market
Weighting schemes for Capitalization % of EAFE GDP
EAFE countries, 2015
United States 50.8% 43.1%
Japan 9.9 9.9
Hong Kong 6.5 0.7
United Kingdom 5.6 6.8
France 4.2 5.8
Germany 3.5 8.1
Canada 3.2 3.7
Switzerland 3.1 1.6
Sweden 2.6 1.2
Australia 2.4 3.2
Spain 1.6 2.9
Netherlands 1.5 1.8
Singapore 1.3 0.7
Italy 1.2 4.4
Belgium 0.8 1.1
Norway 0.4 0.9
Denmark 0.4 0.7
Finland 0.3 0.6
Ireland 0.3 0.6
Austria 0.2 0.9
New Zealand 0.2 0.4
Portugal 0.1 0.5
Greece 0.1 0.5

Source: Authors’ calculations using data from Tables 25.1 and 25.2.

bod77178_ch25_853-[Link] 872 04/18/17 09:00 PM

 Final PDF to printer

eXcel APPLICATIONS: International Portfolios

T his Excel model provides an efficient frontier analysis simi-

lar to that in Chapter 7. In Chapter 7 the frontier was based
on individual securities, whereas this model examines the
Excel Questions
1. Find three points on the efficient frontier corresponding to
three different expected returns. What are the portfolio stan-
returns on international exchange-traded funds and enables dard deviations corresponding to each expected return?
us to analyze the benefits of international diversification. The
Excel spreadsheet is available in Connect or through your 2. Now assume that the correlation between the S&P 500
and the other country indexes is cut in half. Find the new
course instructor.
standard deviations corresponding to each of the three
expected returns. Are they higher or lower? Why?

A B C D E F G H I J
58 Bordered Covariance Matrix for Target Return Portfolio
59 EWD EWH EWI EWJ EWL EWP EWW SP 500
60 Weights 0.00 0.00 0.08 0.38 0.02 0.00 0.00 0.52
61 0.0000 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
62 0.0000 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
63 0.0826 0.00 0.00 4.63 3.21 0.55 0.00 0.00 7.69
64 0.3805 0.00 0.00 3.21 98.41 1.82 0.00 0.00 53.79
65 0.0171 0.00 0.00 0.55 1.82 0.14 0.00 0.00 2.09
66 0.0000 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
67 0.0000 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
68 0.5198 0.00 0.00 7.69 53.79 2.09 0.00 0.00 79.90
69 1.0000 0.00 0.00 16.07 157.23 4.59 0.00 0.00 143.47
70
71 Port Via 321.36
72 Port S.D. 17.93
73 Port Mean 12.00
74
75
76 Weights
77 Mean St. Dev EWD EWH EWI EWJ EWL EWP EWW SP 500
78 6 21.89 0.02 0.00 0.00 0.71 0.00 0.02 0.00 0.26
79 9 19.66 0.02 0.00 0.02 0.53 0.02 0.00 0.00 0.41
80 12 17.93 0.00 0.00 0.08 0.38 0.02 0.00 0.00 0.52
81 15 16.81 0.00 0.00 0.14 0.22 0.02 0.00 0.00 0.62
82 18 16.46 0.00 0.00 0.19 0.07 0.02 0.00 0.00 0.73
83 21 17.37 0.00 0.00 0.40 0.00 0.00 0.00 0.00 0.60
84 24 21.19 0.00 0.00 0.72 0.00 0.00 0.00 0.00 0.28
85 27 26.05 0.00 0.00 1.00 0.00 0.00 0.00 0.00 0.00
86
87

economy than the value of its outstanding stocks. Others have even suggested weights
proportional to the import share of various countries. The argument is that investors who
wish to hedge the price of imported goods might choose to hold securities in foreign firms
in proportion to the goods imported from those countries.

Performance Attribution
We can measure the contribution of each of the following decisions to portfolio perfor-
mance following a procedure similar to the performance attribution techniques introduced
in Chapter 24.
1. Currency selection measures the contribution to total portfolio performance attrib-
utable to exchange rate fluctuations relative to the investor’s benchmark currency,
which we will take to be the U.S. dollar. We might use a benchmark like the EAFE
index to compare a portfolio’s currency selection for a particular period to that of a

873

bod77178_ch25_853-[Link] 873 04/18/17 09:00 PM

 Final PDF to printer

874 PART VII Applied Portfolio Management

passive benchmark. The benchmark for currency selection would be the weighted
average of the appreciation of the currencies represented in the EAFE portfolio
using as weights the fraction of that portfolio invested in each currency.
2. Country selection measures the contribution to portfolio performance attributable
to investing in the better-performing stock markets of the world. It can be assessed
from the weighted average of the equity index returns of each country using as
weights the share of the manager’s portfolio in each country. We use index returns
to abstract from the effect of security selection within countries. To measure the
contribution of country selection, we compare the manager’s weighted average to
the weighted average of a benchmark passive allocation, for example, the share of
the EAFE portfolio in each country.
3. Stock selection ability may, as in Chapter 24, be measured as the weighted average
of equity returns in excess of the equity index in each country. Here, we would use
local currency returns and use as weights the investments in each country.
4. Cash/bond selection may be measured as the excess return derived from weighting
bonds and bills differently from some benchmark weights.
Table 25.14 provides an example of how to measure the contribution of the decisions an
international portfolio manager might make.

Table 25.14
Currency
Example of EAFE Return on Appreciation Manager’s Manager’s
performance Weight Equity Index E1/E0 − 1 Weight Return
attribution:
International Europe 0.30 10% 10% 0.35 8%
Australasia 0.10 5 −10 0.10 7
Far East 0.60 15 30 0.55 18

Overall performance (dollar return = local return + currency appreciation)

EAFE: 0.30(10 + 10) + 0.10(5 − 10) + 0.60(15 + 30) = 32.5%
Manager: 0.35(8 + 10) + 0.10(7 − 10) + 0.55(18 + 30) = 32.4%
Loss of 0.10% relative to EAFE

Currency selection
EAFE: (0.30 × 10%) + (0.10 × (−10%)) + (0.60 × 30%) = 20% appreciation
Manager: (0.35 × 10%) + (0.10 × (−10%)) + (0.55 × 30%) = 19% appreciation
Loss of 1% relative to EAFE

Country selection
EAFE: (0.30 × 10%) + (0.10 × 5%) + (0.60 × 15%) = 12.5%
Manager: (0.35 × 10%) + (0.10 × 5%) + (0.55 × 15%) = 12.25%
Loss of 0.25% relative to EAFE

Stock selection
(8% − 10%)0.35 + (7% − 5%)0.10 + (18% − 15%)0.55 = 1.15%
Contribution of 1.15% relative to EAFE

Sum of attributions (equal to overall performance)

Currency (−1%) + country (−.25%) + selection (1.15%) = −0.10%

bod77178_ch25_853-[Link] 874 04/18/17 09:00 PM

 Final PDF to printer

CHAPTER 25 International Diversification 875

Concept Check 25.3

Using the data in Table 25.14, compute the manager’s country and currency selection if portfolio weights had
been 40% in Europe, 20% in Australasia, and 40% in the Far East.

1. U.S. assets are only a part of the world portfolio. International capital markets offer important SUMMARY
opportunities for portfolio diversification with enhanced risk–return characteristics.
2. Exchange rate risk imparts an extra source of uncertainty to investments denominated in foreign
currencies. Much of that risk can be hedged in foreign exchange futures or forward markets, but a
perfect hedge is not feasible unless the foreign currency rate of return is known.
3. Returns in different countries are far from perfectly correlated. Therefore, there is a benefit from
international diversification. The minimum variance global portfolio has considerably lower
volatility than almost any individual country index, including that of the U.S. More importantly,
single-country stock indexes, again including those of the U.S., plot considerably inside the effi-
cient frontier constructed when foreign equity markets are added to the investment menu. There-
fore, international investing offers ample opportunities to improve the risk-reward trade-off.
4. International investing entails an added dimension of political risk, including uncertainty about
government and social stability, democratic accountability, macroeconomic conditions, interna-
tional trade, and legal protections afforded individuals, businesses, and investors. Several services
now exist that sell information about political risk to interested parties.
5. Several world market indexes can form a basis for passive international investing. Active interna-
tional management can be partitioned into currency selection, country selection, stock selection,
and cash/bond selection.

exchange rate risk political risk country selection KEY TERMS

interest rate parity relationship Europe, Australasia, Far East stock selection
covered interest arbitrage (EAFE) index cash/bond selection
relationship currency selection

Interest rate parity (covered interest arbitrage) for direct ($/foreign currency) exchange rates: KEY EQUATIONS
1 + ​r​  f​​  (U.S.)
​​F​  0​​ = ​E​  0​​ ​ ___________
  
    ​​
1 + ​r​  f​​  (foreign)
Interest rate parity for indirect (foreign currency/$) exchange rates:
1 + ​r​  f​​  (foreign)
​​F​  0​​ = ​E​  0​​ ​ ___________
  
    ​​
1 + r​ ​  f​​  (U.S.)

1. Do you agree with the following claim? “U.S. companies with global operations can give you PROBLEM SETS
international diversification.” Think about both business risk and foreign exchange risk.
2. In Figure 25.2, we provide stock market returns in both local and dollar-denominated terms.
Which of these is more relevant? What does this have to do with whether the foreign exchange
risk of an investment has been hedged?

bod77178_ch25_853-[Link] 875 04/18/17 09:00 PM

 Final PDF to printer

876 PART VII Applied Portfolio Management

3. Suppose a U.S. investor wishes to invest in a British firm currently selling for £40 per share. The
investor has $10,000 to invest, and the current exchange rate is $2/£.
a. How many shares can the investor purchase?
b. Fill in the table below for rates of return after one year in each of the nine scenarios (three
possible share prices denominated in pounds times three possible exchange rates).

Dollar-Denominated Return
for Year-End Exchange Rate
Pound-Denominated
Price per Share (£) Return (%) $1.80/£ $2/£ $2.20/£

£35
£40
£45

c. When is the dollar-denominated return equal to the pound-denominated return?

4. If each of the nine outcomes in Problem 3 is equally likely, find the standard deviation of both the
pound- and dollar-denominated rates of return.
5. Now suppose the investor in Problem 3 also sells forward £5,000 at a forward exchange rate of
$2.10/£.
a. Recalculate the dollar-denominated returns for each scenario.
b. What happens to the standard deviation of the dollar-denominated return? Compare it to both
its old value and the standard deviation of the pound-denominated return.
6. Calculate the contribution to total performance from currency, country, and stock selection for the
manager in the example below. All exchange rates are expressed as units of foreign currency that
can be purchased with 1 U.S. dollar.

Return on Manager’s Manager’s

EAFE Weight Equity Index E1/E0 Weight Return

Europe 0.30 20% 0.9 0.35 18%

Australasia 0.10 15 1.0 0.15 20
Far East 0.60 25 1.1 0.50 20

7. If the current exchange rate is $1.35/£, the 1-year forward exchange rate is $1.45/£, and the inter-
est rate on British government bills is 3% per year, what risk-free dollar-denominated return can
be locked in by investing in the British bills?
8. If you were to invest $10,000 in the British bills of Problem 7, how would you lock in the dollar-
denominated return?
9. Much of this chapter was written from the perspective of a U.S. investor. But suppose you are
advising an investor living in a small country (choose one to be concrete). How might the lessons
of this chapter need to be modified for such an investor?

1. You are a U.S. investor who purchased British securities for £2,000 one year ago when the
British pound cost U.S.$1.50. What is your total return (based on U.S. dollars) if the value of
the securities is now £2,400 and the pound is worth $1.45? No dividends or interest were paid
during this period.

bod77178_ch25_853-[Link] 876 04/18/17 09:00 PM

 Final PDF to printer

CHAPTER 25 International Diversification 877

2. The correlation coefficient between the returns on a broad index of U.S. stocks and the
returns on indexes of the stocks of other industrialized countries is mostly _____, and the
correlation coefficient between the returns on various diversified portfolios of U.S. stocks is
mostly _____.
a. less than .8; greater than .8.
b. greater than .8; less than .8.
c. less than 0; greater than 0.
d. greater than 0; less than 0.
3. An investor in the common stock of companies in a foreign country may wish to hedge against
the _____ of the investor’s home currency and can do so by _____ the foreign currency in the
forward market.
a. depreciation; selling.
b. appreciation; purchasing.
c. appreciation; selling.
d. depreciation; purchasing.
4. John Irish, CFA, is an independent investment adviser who is assisting Alfred Darwin, the head of
the Investment Committee of General Technology Corporation, to establish a new pension fund.
Darwin asks Irish about international equities and whether the Investment Committee should
consider them as an additional asset for the pension fund.
a. Explain the rationale for including international equities in General’s equity portfolio. Identify
and describe three relevant considerations in formulating your answer.
b. List three possible arguments against international equity investment and briefly discuss the
significance of each.
c. To illustrate several aspects of the performance of international securities over time,
Irish shows Darwin the accompanying graph of investment results experienced by a U.S.
pension fund in the recent past. Compare the performance of the U.S. dollar and non-
U.S. dollar equity and fixed-income asset categories, and explain the significance of the
result of the account performance index relative to the results of the four individual asset
class indexes.

5
Real Returns (%)

4
Account Performance Index
EAFE Index
3 Non-U.S. $ Bonds
U.S. $ Bonds
2 S&P Index

1
Variability
(standard
0 10 20 30 40 deviation)
Annualized Historical Performance Data
(%)

5. You are a U.S. investor considering purchase of one of the following securities. Assume that the
currency risk of the Canadian government bond will be hedged, and the 6-month discount on
Canadian dollar forward contracts is −.75% versus the U.S. dollar.

bod77178_ch25_853-[Link] 877 04/18/17 09:00 PM

 Final PDF to printer

878 PART VII Applied Portfolio Management

Bond Maturity Coupon Price

U.S. government 6 months 6.50% 100.00

Canadian government 6 months 7.50% 100.00

Calculate the expected price change required in the Canadian government bond that would result
in the two bonds having equal total returns in U.S. dollars over a 6-month horizon. Assume that
the yield on the U.S. bond is expected to remain unchanged.
6. A global manager plans to invest $1 million in U.S. government cash equivalents for the next
90 days. However, she is also authorized to use non-U.S. government cash equivalents, as long as
the currency risk is hedged to U.S. dollars using forward currency contracts.
a. What rate of return will the manager earn if she invests in money market instruments in either
Canada or Japan and hedges the dollar value of her investment? Use the data in the following
tables.
b. What must be the approximate value of the 90-day interest rate available on U.S. government
securities?

Interest Rates (APR): 90-Day Cash Equivalents

Japanese government 2.52%
Canadian government 6.74%
Exchange Rates: U.S. Dollars per Unit of Foreign Currency
Spot 90-Day Forward
Japanese yen 0.0119 0.0120
Canadian dollar 0.7284 0.7269

7. The Windsor Foundation, a U.S.-based, not-for-profit charitable organization, has a diversified

investment portfolio of $100 million. Windsor’s board of directors is considering an initial invest-
ment in emerging market equities. Robert Houston, treasurer of the foundation, has made the
following comments:
a. “For an investor holding only developed market equities, the existence of stable emerging
market currencies is one of several preconditions necessary for that investor to realize strong
emerging market performance.”
b. “Local currency depreciation against the dollar has been a frequent occurrence for U.S. inves-
tors in emerging markets. U.S. investors have consistently seen large percentages of their
returns erased by currency depreciation. This is true even for long-term investors.”
c. “Historically, the addition of emerging market stocks to a U.S. equity portfolio such as the
S&P 500 index has reduced volatility; volatility has also been reduced when emerging market
stocks are combined with an international portfolio such as the MSCI EAFE index.”
Discuss whether each of Houston’s comments is correct or incorrect.
8. After much research on the developing economy and capital markets of the country of Otunia,
your firm, GAC, has decided to include an investment in the Otunia stock market in its Emerging
Markets Commingled Fund. However, GAC has not yet decided whether to invest actively or by
indexing. Your opinion on the active versus indexing decision has been solicited. The following
is a summary of the research findings:

Otunia’s economy is fairly well-diversified across agricultural and natural resources,

manufacturing (both consumer and durable goods), and a growing finance sector.
Transaction costs in securities markets are relatively large in Otunia because of high
commissions and government “stamp taxes” on securities trades. Accounting stan-
dards and disclosure regulations are quite detailed, resulting in wide public availability
of reliable information about companies’ financial performance.

bod77178_ch25_853-[Link] 878 04/18/17 09:00 PM

 Final PDF to printer

CHAPTER 25 International Diversification 879

Capital flows into and out of Otunia, and foreign ownership of Otunia securities,
is strictly regulated by an agency of the national government. The settlement proce-
dures under these ownership rules often cause long delays in settling trades made
by nonresidents. Senior finance officials in the government are working to deregulate
capital flows and foreign ownership, but GAC’s political consultant believes that isola-
tionist sentiment may prevent much real progress in the short run.

a. Briefly discuss aspects of the Otunia environment that favor investing actively, and aspects
that favor indexing.
b. Recommend whether GAC should invest in Otunia actively or by indexing. Justify your rec-
ommendation based on the factors identified in part (a).

E-INVESTMENTS EXERCISES
A common misconception is that investors can earn excess returns by investing in foreign bonds
with higher interest rates than are available in the U.S. Interest rate parity implies that any such
interest rate differentials will be offset by premiums or discounts in the forward or futures market
for foreign currency.
Interest rates on government bonds in the U.S., U.K., Japan, Germany, and Australia can be
found at [Link]/markets/rates/[Link].
Spot exchange rates on international currencies can be found at [Link]/
markets/currencies/[Link].
Forward exchange rates on currency futures contracts can be found at [Link]/
trading/fx/[Link].

1. Select one of these countries and record the yield on a short-term government security from
the Bloomberg Web site. Also make note of the U.S. Treasury yield on an instrument with the
same (or closest possible) maturity.
2. Record the spot exchange rate from the Bloomberg site and the futures contract exchange
rate from the CME Web site for the date closest to the maturity of the investment you chose in
the previous question.
3. Calculate the rate of return available on the foreign government security, converting the for-
eign currency transactions into dollars at the current and forward exchange rates.
4. How well does interest rate parity seem to hold? Are there bargains to be found in other cur-
rencies? What factors might account for interest rate parity violation?

SOLUTIONS TO CONCEPT CHECKS

1. 1 + r (US) = [1 + rf (UK)] × (E1/E0)
a. 1 + r (US) = 1.1 × 1.0 = 1.10. Therefore, r (US) = 10%.
b. 1 + r (US) = 1.1 × 1.1 = 1.21. Therefore, r (US) = 21%.
2. You must sell forward the number of pounds you will end up with at the end of the year. This value
cannot be known with certainty, however, unless the rate of return of the pound-denominated
investment is known.
a. 10,000 × 1.20 = 12,000 pounds.
b. 10,000 × 1.30 = 13,000 pounds.

bod77178_ch25_853-[Link] 879 04/18/17 09:00 PM

 Final PDF to printer

880 PART VII Applied Portfolio Management

3. Country selection:

​(0.40 × 10%) + (0.20 × 5%) + (0.40 × 15%) = 11%​

This is a loss of 1.5% (11% versus 12.5%) relative to the EAFE passive benchmark.
Currency selection:

​(0.40 × 10%) + (0.20 × (−10%)) + (0.40 × 30%) = 14%​

This is a loss of 6% (14% versus 20%) relative to the EAFE benchmark.

bod77178_ch25_853-[Link] 880 04/18/17 09:00 PM