Today’s conventional wisdom suggests otherwise. In recent years, pension and endowment funds have gradually built their holdings of foreign equities to significant levels, often at the expense of their U.S. equity positions. More than 1,000 U.S.-operated equity mutual funds—one out of every four—invest primarily in international equities, and even our largely domestic funds have nearly 7 percent of their assets invested in foreign issues. The global investing strategy is a favorite of academic theorists, some of whom recommend that the ideal equity portfolio should consist of holdings of each nation’s stocks at their world-market weight. As of mid-1998, such a strategy would have yielded the following weightings:

Nation	Percent
United States	48%
Japan	9
United Kingdom	10
Germany	5
France	4
Emerging markets	4
Other	20
Total	100%

Note that more than half of the assets of this “ideal” portfolio for a U.S. investor would be placed outside the United States.

This strategy rests primarily on a simple premise: Because foreign markets have experienced patterns of volatility that are different from those of the U.S. market (and, to an important degree, different from one another), their inclusion in a portfolio of U.S. equities would reduce the portfolio’s total volatility and hence provide higher risk-adjusted returns. Other arguments advanced in favor of the strategy include: the enhanced stability inherent in a more diversified economic base; higher potential growth rates; and cheaper valuations in world markets. But these are speculative arguments that may or may not prove valid.

The other side of the story was well presented, in late 1997, by Wall Street Journal columnist Roger Lowenstein.¹ Describing the strategy as “Global-Investing Bunk,” he wrote that “this faddish bit of investment wisdom was exposed as nonwisdom in 1997.” Citing the collapse in Asian stocks, he challenged the notion that “the ‘sound’ investor is one who has moved a goodly chunk of his money out of the society he knows to countries with which he is unfamiliar, each according to its market weights.” Lowenstein concluded that “while gambling in every country is wiser than gambling in one, not gambling at all is wiser still.”

TEN YEARS LATER

The Global Portfolio

Today’s global portfolio looks surprisingly similar to its profile of a decade ago. The weight of U.S. stocks has declined slightly, from 48 percent to 44 percent, and the developed countries have more or less held their own. Only the emerging markets have shone, rising from 4 percent in 1998 to 9 percent a decade-plus later.

Nation	Percent
United States	44%
Japan	11
United Kingdom	8
Germany	4
France	4
Emerging markets	9
Other	20
Total	100%

Currency Risk—and Returns

I, too, have serious reservations about a full market-weighted global strategy. It involves a very heavy layer of one particular risk that an equity investor never need assume: currency risk. Returns earned from investing in stock valued in one’s home currency are measured in the coin of the realm in which the investor earns, spends, and saves (sometimes to your advantage as a citizen, to be sure, sometimes not). Even if foreign investments were to provide the same rate of return—measured in their local currencies—as U.S. investments, the returns earned might be very different, depending on the strength or weakness of the dollar in world markets. A strong dollar reduces the returns earned by U.S. investors in foreign markets; a weak dollar increases the returns earned in foreign markets.

In the very long run, because the mechanics of government financing and international trade should equalize currency values, foreign currency risk relative to the U.S. dollar should be a neutral factor in global markets. But over interim periods, the road to parity may be rocky. For example, during the 30-year history of the Morgan Stanley Capital International Europe, Australasia, Far East (EAFE) Index, the U.S. dollar was weak from 1970 through 1980, strong through 1984, and very weak until 1995. Since then, it has rallied nicely. On balance, however, the dollar in 1998 is worth roughly what it was worth in 1980. These fluctuations led to considerable misunderstanding about the returns earned in international markets. A fine example came during the decade that ended in 1994. The upper section of Table 8.1 shows that foreign returns for U.S. investors (i.e., in dollar terms) were far higher than returns on U.S. stocks during that decade.

Given this scenario, investors could easily have been (and probably were) led to believe that foreign markets had outpaced the U.S. market and therefore promised better future growth. The 3.6 percent enhancement in annual return—over a very strong U.S. market—had led to 137 percentage points of extra capital accumulation over a decade. Why not jump on the bandwagon?

Why not, indeed? Because the excess returns in international markets were, to an important degree, an illusion. Although investors had earned a handsome return in the currency markets, their returns from foreign stock markets were, in fact, rather ordinary. Measured in local currencies, international market annual returns lagged the U.S. market by 4.1 percentage points annually, just a bit more than the margin by which they appeared to lead it. The inordinate weakness in the dollar had more than doubled the cumulative international return, as shown in the lower section of Table 8.1.

I believe that the performance of foreign stocks for U.S. investors, in the long run, will be determined by each nation’s fundamental returns (based on dividend yields and earnings growth), rather than by currency returns. Taking into account the national economic growth rates around the globe during the 1990s—and comparing them to the powerful global reach, entrepreneurial energy, and technology leadership we see in our nation today—it would be logical to expect U.S. growth to exceed the growth in other nations.

TABLE 8.1 Global Returns for the 10 Years Ended December 31, 1994

TEN YEARS LATER TABLE 8.1 Global Returns for the Period 12/94 through 6/09

On the other hand, the United States may well be at the pinnacle of its economic cycle, while many of the European powers, Japan, and the emerging markets—beset, respectively, with high unemployment, overextended financial institutions, and deep recessions—may have reached some sort of nadir. Further complicating the matter, it is never clear whether these economic factors are accurately reflected in present market valuations. The race for superior market returns around the globe is never an easy one to call.

But there was what seemed to me an easy call in 1994. As it turned out, that call has proven correct so far: such weakness in the dollar could not continue indefinitely. Given that likelihood, past performance data so heavily influenced by the weakness in the dollar were deeply flawed. Looking to the future without being aware that the weak dollar had added 7.7 percentage points annually to foreign stock returns would result in highly exaggerated expectations. In fact, from the close of 1994 through mid-1998, the dollar’s strength was restored and actually reduced foreign returns by an annualized rate of 4.6 percentage points.

TEN YEARS LATER

Currency Risks and Returns

Fifteen years have now passed since my self-described “easy call” that the “weakness in the dollar [from 1984 through 1994] could not continue indefinitely.” The call proved to be right on the mark. The dollar quickly strengthened through 2001, and the long back- and-forth cycle continued as the dollar weakened in 2002-2008, only to strengthen a bit in mid-2009, before weakening again. For the full period, the dollar declined 15 percent against the Federal Reserve’s major currency index. The dollar’s weakness during this long span, from 1994 to mid-2009, produced only a slightly higher return for international stocks in dollar terms (3.7 percent) than for those measured in local currencies (3.5 percent). Nonetheless, the returns of both lagged the return of the Standard & Poor’s 500 Index (6.8 percent) for the full period.

The Global Efficient Frontier

Many investors reject the full market-weighted global strategy but endorse a more sophisticated form of analysis that sets the structure of the global portfolio. The analysis involves the calculation of an efficient frontier, which is designed to determine the precise allocation of assets between U.S. and foreign holdings. The goal is a combination that promises the highest return at the lowest level of risk (i.e., the lowest volatility of return acceptable to the investor). I am skeptical of this approach as well, for the efficient frontier is based almost entirely on past returns and past risk patterns. That bias may be unavoidable—after all, history is our only source of hard statistics—but past relative returns of stock portfolios and (albeit to a much lesser degree) past relative volatility are not always harbingers of the future, and may even be counterproductive.

Consider, for example, the efficient frontier that would have been drawn at the end of 1988, the high-water mark for relative returns in international markets. As shown in the upper section of Table 8.2, the optimal combination of the highest return for the lowest risk—and, for aficionados of the theory, the requisite asset allocation—would have called for 50 percent invested in foreign stocks and 50 percent in U.S. stocks. How would it have worked out? Not particularly well. Indeed, after 1988, that portfolio proved to be rather insufficient, if not decidedly inefficient. With an annual return of 11.4 percent (compared to 19.2 percent in the United States), not only did it sharply lag the return of the all-U.S. portfolio, it also actually proved subject to slightly higher risk than the U.S. portfolio (standard deviation of 12.6 percent versus 12.2 percent). That combination of return and risk is hardly a winning combination.

TABLE 8.2 The Efficient Frontier

Today, an investor putting money to work using this theory would settle on a rather different global portfolio. Based on history, the efficient frontier is backward-looking. So the investor currently seeking maximum return and minimum risk would select an efficient portfolio based on the returns of the respective markets over the past decade, when the results were as shown in the lower section of Table 8.2.

Given a decade of high returns and low volatility in the United States—and the reverse in the rest of the world—U.S. stocks now represent 80 percent of the portfolio, nearly double their weight a decade earlier. Will this new global allocation lead the followers of the efficient frontier theory in the right direction? We have no way of knowing. But experience leads us to conclude that it is rather unlikely to provide the optimal answer.

Another problem with the efficient frontier theory, it seems to me, is that extremely small variations in risk may separate the optimal portfolio from those deemed less efficient. For example, in the decade ending in 1988, the standard deviation of an 80/20 U.S./international mix was 15.1 percent, compared to 14.4 percent for a 50/50 mix (purportedly most efficient). That difference—less than a single percentage point—is so small as to be almost invisible to any real-world investor, particularly one who is not willing or able to engage in the arcane methodology required for calculating standard deviations of monthly return, leaving aside whether such deviations are a valid proxy for risk.

Based on the nine and a half years ended in June 1998, the central tolerances are equally minuscule. For example, the 11.8 percent standard deviation for the hindsight-based 80/20 U.S./international efficient portfolio with the lowest risk would have offered a reduction of less than one percentage point below the 12.2 percent figure for a portfolio holding 100 percent in U.S. equities. For intelligent investors to allow their entire portfolio strategy to be based on these truly trivial past differences in risk—really an elusive proxy for risk—seems a wholly unwarranted triumph of process over judgment.

Contrasting the periods ending in 1988 and in June 1998, Figure 8.1 crystallizes two important realities of the global efficient frontier: (1) vast shifts in the frontier may take place over a decade, and (2) variations in risk near the efficient point of each curve are inconsequential, despite large variations in asset allocation. Slavish reliance on history seems particularly flawed in markets where currency fluctuations create substantial extra risk.

There is, in the final analysis, only one risk that equity investors need to assume: market risk—the inevitable truth that all stock portfolios fluctuate in value. For better or worse, most investors choose to assume two additional risks: style risk, or choosing mutual funds and stock portfolios with a particular bias, such as those focused on large - cap value stocks, or small-cap growth stocks, or any other stocks whose returns are expected to vary from the total stock market over time; and manager risk, or selecting a mutual fund whose portfolio manager may or may not provide the optimal portfolio within the fund’s style category.

FIGURE 8.1 Which Efficient Frontier? U.S. versus International Holdings

Leaving aside for the moment the wisdom of assuming those two extra risks—which nearly all investors take for granted—I see no reason for investors to assume yet a third extra risk: currency risk. But, to those who, in their own wisdom and judgment, accept the thesis that global investing is necessary, I reaffirm my rule -of-thumb recommendation: Limit international holdings to no more than one-fifth of the equity portfolio.

TEN YEARS LATER

The Global Efficient Frontier

As shown in Table 8.2, the essentially zero returns earned in both U.S. and international stocks during the recent decade were a far, far cry from the mostly double-digit returns of the magical—and clearly unrepeatable—preceding two decades. By most measures, however, volatility risks were little changed.

New Figure 8.1 clearly reaffirms my skepticism about relying on the inevitably backward-looking efficient frontier concept to set investment strategy. As I wrote then, “experience leads us to conclude that [the 1998 efficient frontier] is rather unlikely to provide the optimal answer.” A glance at the new chart shows that the sweeping variables of the efficient frontier of the 1980s and then the 1990s have been replaced by a cramped arc in which the difference in annual returns between 100 percent U.S. stocks and 100 percent international stocks is trivial. What will the efficient frontier look like a decade hence? A guess: significantly different from its pattern in any of the three past decades. And that is why charting a supposedly efficient frontier, however interesting as an historical artifact, is a weak reed on which to lean.

International Economies and Financial Markets

So far, I have not dealt with one of the most obvious risks of putting money to work abroad: the contrasts between the governments, economies, and financial systems of most foreign countries and those of the United States. As a theoretical matter, those risks tend to be subsumed by the marvelous arbitrage pricing mechanism that is implicit in the financial markets. Through this mechanism, market prices are assumed to take into account all existing information about a stock—including all stocks of all nations—and, through the decisions of informed buyers and sellers, to reach a price that accurately balances potential risk against prospective reward.

In retrospect, the record clearly shows that huge shifts take place in the relative positioning of national economies—sometimes over an extended period of time, and sometimes with remarkable speed. The financial market is a stern taskmaster that brooks few compromises. Consider Japan. A decade ago, U.S. investors were worried that the Japanese economy, “the rising sun,” would eventually dominate the world. In Tokyo, soaring stock prices escalated the total capitalization of Japanese stocks to nearly half of the world’s entire market capitalization. Japan’s dominant 43 percent share was half again the size of the U.S. market’s 28 percent portion at the time.

Since then, Japan’s economy has withered while most other economies (especially that of the United States) have flourished. The Japanese government’s fiscal and monetary policies have seemed to hurt rather than help that nation’s economy, and its banking system remains overextended and permeated with problem loans. The Nikkei stock market index has fallen from 27,700 yen to 15,800 yen in mid-1998, about half its value a decade ago.

Meanwhile, other world markets have prospered. In the United States, during the same period, the Dow Jones Industrial Average increased from 2,140 to 8,950—a 320 percent rise. The European markets (according to the MSCI-EAFE-Europe Index) rose 275 percent in local currency terms. Japan’s astonishing fall from grace was reflected in its global weighting in mid-1998: 9 percent of world markets, or about one-fifth of its 43 percent weighting a decade before. During the same period, the U.S. market has grown to 48 percent of the value of the world’s total capitalization, or more than half again our 28 percent portion in 1988. Note the striking similarity of our world market share in 1998 to Japan’s a decade earlier. Whether that parallel will continue with a comparable erosion in the years to come remains to be seen, but it is worth thinking about. Clearly, if you believe it will happen, you should overweight foreign holdings. Table 8.3 reflects the relative changes in the capitalization of world markets in 1988 and mid-1998. Bear in mind that those dramatic shifts took place during less than a decade—a rather brief span of financial history.

Reversal of Fortune?

In the early autumn of 1998, the total U.S. stock market was up about 3 percent for the year, and international markets were down about 7 percent. Stock markets in Western European nations were higher, but virtually all other global markets had tumbled. Japan was down about 20 percent, the emerging markets were off about 40 percent, and the Russian markets had plummeted almost 80 percent. To add insult to injury, weaknesses in local currencies of most foreign nations constituted a significant portion of the losses suffered by U.S. investors.

This additional enhancement of the relative returns on U.S. stocks can be interpreted in one of two ways: (1) it further validates the “Acres of Diamonds” thesis of this chapter, or (2) it makes the opportunities for U.S. investors abroad even more attractive than they were earlier. After all, there is considerable substance to the argument that returns in international and U.S. markets will revert to the mean of a uniform global return on stocks over the long term, as happened, on balance, from 1960 through 1997.

Even in these days of U.S. ascendancy, who can be absolutely confident that the U.S. dollar, having soared, will not tumble? Or that a rival nation (or block of nations, such as the European Community) will not supplant the United States as the world economic leader? Or that present world market values are now priced in anticipation of the worst outcomes around the world and the best outcome in the United States? The answer is: no one, least of all I. Reversals of fortune in the financial markets have proved to be more the rule than the exception.

But a trap awaits investors who decide the best is yet to be for foreign markets. It is the trap of market timing. Reversals of fortune often come when least expected. It is easy to be too late, or too early, to take advantage of them, as was the case after their big 1997 tumble. Maybe the worst is finally over abroad. But maybe it is not. We just don’t know.

A current example of the risk involved in global investing is Southeast Asia. Through mid -1997, many global investors had looked to these emerging markets as offering an unusually favorable opportunity for earning superior long-term returns. And, during the 1980s and 1990s, both the economies and the markets of Indonesia, South Korea, Malaysia, Singapore, Thailand, and the Philippines had indeed distinguished themselves. In the newly global economy, their populations and economies were growing apace, and soaring stock returns doubled their weight in world markets, from 1.9 percent in 1991 to 3.8 percent as 1997 began. But, by autumn of 1997, their government-dominated financial systems weakened, their currencies plummeted, and their economies slumped.

Rare was the Southeast Asian market that did not tumble by 40 percent or more in local currency terms and another 40 percent in dollar terms. Declines of 80 percent or more in value were the norm for U.S. investors—all in the span of just a few months. With other world markets marching upward, the relative weight of the Southeast Asian markets fell by an astonishing 70 percent. As 1998 began, their weight was 1.2 percent of world markets, or less than one-third of their weight only a year earlier, when they reached the pinnacle of their popularity with fund investors. The problems persisted in 1998, and these emerging markets continued to deteriorate. These reversals have given investors a humbling lesson in the risks of global investing. Those risks are especially high in nations where U.S. standards for accounting, financial transparency, and liquidity have not yet been attained.

TABLE 8.3 Capitalization of World Markets: 1988, 1998, and 2008

TEN YEARS LATER

International Financial Markets

As noted earlier, the ebbing and flowing of economic growth and the changes in investor confidence between overoptimism and overpessimism bring their banes—or their blessings—to economies and markets around the globe. Table 8.3 (updated in this edition) clearly reflects the collapse of the Japanese stock market that began in the late 1980s and the remarkable surge in the U.S. stock market during the 1990s.