Advertisement
Many planners would go along with this basic tenet of asset allocation: It makes sense to combine types of investments that don't all move in the same direction.
That said--or assumed--the reverse corollary is obvious: If international correlations rise, there's less reason to buy foreign securities. Are those correlations increasing, as some contend? "It depends," says Tom Idzorek, research director of Ibbotson Associates in Chicago. "Changing time frames can change the story."
TIMING IS EVERYTHING
For example, Ibbotson ran numbers using monthly rolling five-year returns, matching the S&P 500 with Morgan Stanley's EAFE (Europe, Australasia, Far East) Index, a widely used international benchmark. Starting with a correlation of around 0.6 in late 1974, the correlation dipped as low as 0.25 in late 1996, shot up in late 1998, and gradually rose to around 0.9 in mid-2005. (A 1.0 correlation indicates a perfect match; a negative number shows assets that move in opposite directions.)
"The picture looks different if you go from monthly to annual returns and from five- to 10-year rolling returns," Idzorek says. Starting in late 1979 with correlations at 0.7, the same dip is evident through the mid-1990s, but the recent rise in correlations has been back merely to 0.7, the same as it was 25 years ago.
"The longer the time horizon, the greater the net benefit," he says. "Using a longer time frame, international correlations haven't changed much since 1979." For clients with decades-long investment horizons, spreading dollars around the globe may well reduce correlation.
Still, even the longer-term numbers tracked by Ibbotson show increasing correlation over the last decade, especially between large-cap U.S. stocks and the large-cap stocks of developed foreign countries. That correlation is now back to its highest levels of the past 25 years.
"International correlations may be on the rise, and the absolute value provided by international investing is not quite as high as it has been during some past periods," Idzorek says. "Even so, U.S. and foreign markets are not perfectly correlated, so there is still a diversification benefit."
DOWNSIDE HELP?
That benefit is most apparent when U.S. markets are weak and foreign markets provide better returns. Unfortunately, this isn't always the case. "Some studies have indicated that cross-country correlations are higher in falling markets than in rising markets," says Jim Davis, vice president of research at Dimensional Fund Advisors (DFA) in Santa Monica, Calif. "This is the opposite of what an internationally diversified investor would like to see."
Again, time may be on the side of long-term investors. When Davis did his own study of internationally diversified portfolios using monthly data from 1970 to 2003, those downbeat studies were confirmed: Downside correlations with U.S. equities were higher than upside ones for each international equity class reviewed.
"Changing from monthly returns to quarterly returns made a big difference," Davis says. With the possible exception of Continental Europe, none of the downside correlations with the United States were reliably higher than the upside correlation. Thus, for long-term investors, the DFA study found that international diversification can reduce standard deviation--and decrease portfolio volatility--in falling as well as rising U.S. markets.
"When you measure by longer intervals, there's more of a benefit to international diversification," Davis says. "Correlations may have increased recently, which might somewhat reduce the potential benefits of international investing, but that does not make them go away. Even if correlations are above 0.8, as long as they are not complete, there still can be some benefit to investors."
There may be times, however, when those benefits don't appear to be substantial. Foreign stocks provided little help in the 1973 to 1974 and 2000 to 2002 bear markets nor during the October 1987 market crash, according to data supplied by Ibbotson. For example, a portfolio fully invested in the S&P 500 would have lost 45% from September 2000 to September 2002. With a 30% exposure to EAFE, the loss still would have been 44%, while a 70% S&P 500/25% EAFE/5% emerging markets mix would have been down 43%.
"In times of distress, all financial assets go down," says Weston Wellington, a vice president at DFA. "But you should not conclude from these events that there is no benefit to diversification."
Idzorek concurs that even such a two-year period is too short a time to gauge the value of international diversification. He prefers a decade-by-decade approach. "In both the 1970s and the 1980s, replacing a 100% S&P 500 portfolio with a 75%/25% mix, S&P 500 to EAFE, would have raised returns and substantially reduced standard deviations," he says. "In the 1990s, including international stocks would have lowered returns for U.S. investors but still provided some risk reduction."
More recently, the Ibbotson data from the beginning of 2000 through the first half of 2005 shows that this 75%/25% blend would have had a slightly smaller loss, with lower volatility, than an all-S&P 500 portfolio. However, risk reduction in the current half-decade is down sharply. In the 1980s, adding 25% exposure to EAFE would have dropped the standard deviation from 19.39% to 17.71%; since the start of 2000, moving 25% into EAFE barely dropped the standard deviation, from 15.687% to 15.2%.
Over the entire 351/2 years, the S&P 500 and EAFE had about the same returns--geometric means of 11.1% and 10.9%, respectively--so diversification would have had only a modest impact. But risk reduction was significant, since adding 25% EAFE to an S&P 500 portfolio would have dropped standard deviations from 17.31% to 16.03%.
BE BROAD
In a true crisis, there's a flight to quality. "That's why you have bonds in a portfolio," Idzorek says. Indeed, bonds posted solid returns in 2000 to 2002.
Diversifying via bonds can cross borders, too. For example, Principal Global Investors in Des Moines, Iowa, conducted a correlation study on 18 fixed-income markets over an eight-year period ending in March 2004. Nearly half of the resulting combinations (72 out of 153) had a correlation of less than 30%, with many being close to zero or negative.
"Since then, correlations have risen among fixed-income markets, because global economies have been in sync for the past few years," says Bob Baur, managing director and head of economics and equity trading for Principal Global. "International bond correlations were coming from a low base, though, so they still are low, compared with the equity markets."
Investing in foreign bonds isn't the only way that planners can tailor international positions to attain lower correlations. The three Yale professors quoted above looked at correlations over the past 150 years and discovered that "diversification opportunities among these major markets [the U.S., U.K., Germany, France] have reached a 150-year low." Even during the Great Depression, when correlations previously peaked, "these markets provided greater opportunities for risk spreading than they do today."
Nevertheless, the Bulldog Three did not give up on international exposure. They found diversification benefits from the "increasing number of world markets available," in addition to some benefits from less-than-perfect correlations.
The Ibbotson numbers also show that emerging markets are less correlated than developed foreign markets to the United States (although the data covered a much shorter time period). Using yearly rolling 10-year returns, for example, Ibbotson shows the correlation between the S&P 500 and the S&P/IFC Investable Emerging Markets Index is now a mere 0.2.
Other categories also might merit a place in client portfolios. "A study in the 1990s showed bigger benefits among small-caps than large-caps for international diversification," Davis recalls. "I believe this is still the case."
Solid evidence to support this position comes from Cathy Pareto, senior financial advisor at Investor Solutions in Coconut Grove, Fla., who ran some correlation numbers to help with this article. From 2002 through 2004, she discovered, the correlation between the S&P 500 and DFA's Emerging Markets Index has been 0.83. This brought the eight-year correlation (1997 through 2004) up to 0.7, indicating that recently emerging market and U.S. equities have become much more likely to move together.
Lower recent correlations, though, can be found in small companies: Between the S&P 500 and DFA's International Small Company Portfolio, Pareto found the three-year correlation has been 0.64 and the eight-year correlation was 0.51. Correlations have been particularly low between the S&P 500 and Morgan Stanley's Small-Cap Japan Index--a mere 0.19 over the past three years, which actually dropped the eight-year correlation to 0.23.
"There is still a tremendous benefit to being invested in foreign positions, regardless of creeping correlation," Pareto concludes. "We have half of clients' equity allocations in foreign stocks. Our tendency is to overweight value and small-company stocks, so that's true on the international side as well as on the domestic side."
WHAT ADVISERS THINK
Planners with less exposure to international equities also express a desire to avoid hefty correlations with the U.S. market. "We recommend that clients have 15% to 20% of their equity positions outside the United States," says Margie Carpenter, portfolio manager of McCormack Advisors International in Cleveland. "At times, as much as half of that international equity position has been in emerging markets. Low correlation is one reason we include emerging markets to this extent, in addition to the growth potential."
Carpenter's firm makes an effort to "mix and match" foreign funds, as she puts it. "We use the Lazard International Equity Portfolio, which has lots of large European companies. T. Rowe Price New Asia matches up nicely because it provides exposure to other parts of the world."
Brett Gallagher, co-manager of New York-based Julius Baer Global Equity Fund, cautions planners against relying solely on geographical diversification in their asset allocations. "Some industries are global, with very high correlations among firms no matter where they're based," he says. Thus, if clients invest in international funds that hold many energy, pharmaceutical, technology and auto stocks, they may not be getting a great deal of benefit from non-correlation, even if those firms' headquarters are in different countries.
On the other hand, Gallagher points to some industries that are domestic, with much lower cross-border correlations: utilities, banks, real estate, health care, hotels and restaurants and telecommunications, for example. Such sectors may provide genuine exposure to non-U.S. economies. By looking into funds' holdings, planners can see not only where their firms are coming from but also what they do there.
Senior Editor Donald Jay Korn has been writing about investments and tax and estate planning for Financial Planning since 1985.
