The CPA Journal

The risk/reward tradeoff seems obvious: the greater the return, the higher the volatility. (Of course, there is actually a trade-off of one kind of risk for another: Stocks may be risky because of the potential of large losses in the short run, but cash is risky in that it may not return enough gains to meet long-term needs.)

And yet, return can be increased without increasing volatility, and volatility can be significantly decreased without decreasing return. To appreciate this seeming anomaly, consider the Nasdaq index and the Dow Jones Industrial Average from 1999–2000 (Exhibit 1). The Nasdaq rose 86% in 1999, the Dow 27%. The next year, the Nasdaq fell 39% and the Dow fell a more modest 5%. $10,000 invested in Nasdaq at the beginning of 1999 had ballooned to $18,610 by the end of the year but ended 2000 at $11,320. The same amount invested in the Dow rose to “only” $12,730 in 1999, but ended 2000 at $12,130. Higher volatility did not reap a higher return in this case.

The simple reason for this is that gains are easier to demolish than losses are to make up. A 50% loss will wipe out a 100% gain: On the other hand, a drop of 50%, requires a 100% gain to make up. In this instance, Nasdaq’s 39% decline in 2000 wiped out almost all of 1999’s 86% gain. The Dow’s 5% drop in 2000 left intact more of 1999’s 27% gain. No matter how many thousands of percent a stock may rise (and some stocks did rise that much in the late ’90s), a subsequent drop of 90–99% (which many such stocks also experienced) can wipe out an investment position.

This also explains why risk and reward cannot be judged solely by averages. After all, the Nasdaq’s return averaged 24% annually over those two years, whereas the Dow’s return averaged only 11%. A proper assessment of any investment must include not only average return, but also volatility. Too much of the latter can destroy the benefit of the former.

Combining some of both the Nasdaq and the Dow in one portfolio would have captured some of the 86% gain in 1999 and benefited from the gentler loss of 5% in 2000. An investment split equally between the Nasdaq and the Dow, rebalanced annually, would have had a higher total return than either index alone.

The reason for this premium is the very different paths the two markets took over those two years. When one was doing well, the other was not. Rebalancing at the end of the first year (bringing the proportion of the two investments back to 50/50) would have locked in the Nasdaq’s gains. Rebalancing puts volatility to work by forcing a high sale and low purchase. Combining just two assets in one portfolio—the simplest example of diversification—resulted in reduced volatility and increased return.

Low Correlation

The driver of diversification is low correlation, meaning that two investments perform substantially differently under the same circumstances. In the above example, it was the very different performances of the Nasdaq and the Dow. Without low correlation, combining them would have done little good.

The author’s research team analyzed the performance of major asset classes over the past two decades [U.S. growth and value stocks, stocks of developed and developing foreign markets, intermediate bonds, and real estate investment trusts (REIT)], using all benchmark data. The analysis confirmed the great truth of the capital markets: Leadership changes over time, and all one can do is guess where the best returns will come from next. In the early 1980s, value stocks came out on top, followed closely by real estate. Subsequently, the major foreign markets sprang to the top, boosted by currency movements. At the end of the 1990s, growth stocks soared and then crashed.

Exhibit 2 depicts these relationships numerically, comparing the returns of U.S. stocks to the other asset classes. The highest possible correlation is 1, which signifies two investments that have performed exactly alike. A correlation of 0 indicates no relationship between two investments’ performance. A negative number indicates that two investments move in opposite ways. Thus, the lower the correlation of investments, the greater the advantage of diversification. Growth stocks are most closely correlated to the U.S. stock and market as a whole, followed by value stocks and major foreign markets. Next are REITs and emerging markets, then bonds, and, at the very bottom, cash equivalents. Therefore, the best financial assets to pair with stocks would be money funds or T-bills.

While some cash in a portfolio reduces risk, it also greatly reduces return. A 50/50 mix of cash and stocks returns almost two percentage points less than an all-stock portfolio. Since 1926 (when data is first available), money funds averaged a paltry return of 3.9%, which could be too little to meet an investor’s needs.

Bonds, however, have generally outperformed cash. In fact, while a million dollars invested in cash in 1960 would have grown to $11 million after 40 years, the same amount in bonds would have grown to $19 million. Because stocks and bonds are more highly correlated than stocks and cash, the diversification benefit of the combination should be smaller; however, long-term growth depends upon three factors—correlation, volatility, and return—and the higher average return for bonds more than outweighs the smaller diversification benefit.

Additionally, the low correlation of stocks and bonds makes bonds an excellent “anchor to windward”: Bonds generally do well when stocks do not. In 11 of the 12 years since 1950 in which U.S. stocks declined, the bond market made money. In the eight bear markets of the past 30 years, bonds always made money, and sometimes lots of it (in the most recent, it was a 20% gain versus a 29% loss).

Interestingly, a 100% bond portfolio is actually slightly more volatile than one with some stocks in it. Bonds do have an occasional losing year. Because of their low correlation to bonds, riskier investments like stocks can actually reduce the overall risk of a portfolio.

The Building Blocks

Stocks and bonds are thus a good foundation on which to build a portfolio, but what is the appropriate mix? While asset allocation will vary for each investor, the classic stock/bond combination is 60/40. Favored by institutional investors because of its superb risk/return tradeoff, this mix has enough stocks for long-term growth without unproductive volatility and enough bonds for an anchor to windward without falling behind inflation.

REITs, whose returns historically have not been highly correlated with stocks and bonds, are also excellent for diversification. In fact, their correlations have been declining over the past decade. Apportioning 20% of an all-stock portfolio to REITs has reduced volatility more than return; adding the same proportion of REITs to an all-bond portfolio has increased return without significantly increasing volatility. In a fully diversified portfolio, setting aside a 10% portion for REITs lowered volatility while maintaining return.

This hypothetical portfolio now contains 55% stocks, 35% bonds, and 10% REITs, but which kinds of stocks will provide true diversification? The author’s research team has concluded that the kind of stocks chosen for a portfolio is more important than how many, as long as there are not too few or too many. The fewer stocks in a portfolio, the more volatile it is.

Are index funds the answer? Mimicking the market can ensure that volatility will not be any higher than the market’s, but then neither will the return. The goal is to reduce volatility and increase return, which can be accomplished by identifying stocks that will not only beat the market as a group in the long run, but also perform differently from one another in the short run.

Investing with Style

Growth and value stocks are good candidates. Because they flourish under different economic and market conditions, they tend to alternate performance leadership (i.e., they are poorly correlated). The author’s research shows that a portfolio allocated 50/50 to growth and value stocks, regularly rebalanced, would have returned 15.4% from 1981 to 2001, compared to the S&P 500’s 14.2%; this at a volatility of 15.1%, compared to the S&P 500’s 15.8%. A $100,000 investment in the S&P 500 would have grown to $1.6 million, but the same amount evenly invested in growth and value stocks would have grown to $2.0 million, with less risk over the entire period.

Applying the same principle to the hypothetical portfolio above and regularly rebalancing over the past two decades would produce nearly the same return as stocks alone (14.2% for stocks, 13.5% for the portfolio), but with far less volatility (15.8% and 9.9%, respectively). For many investors, that smoother ride means greater peace of mind, and greater likelihood they will have money when they need it.

Overseas Markets

Although equities have earned similar returns regardless of where in the world they trade, markets around the world do not perform in lockstep, so their correlation with U.S. stocks ranges from imperfect to weak. The author’s research team has found the best risk/reward tradeoff to be roughly 70% of a stock portfolio in U.S. securities, 25% in major foreign markets, and 5% in emerging markets. Although the U.S. allocation is larger than the proportionate size of its economy, if most of an investor’s earnings and expenses are denominated in U.S. dollars and the investor is subject to U.S. laws, it seems prudent that most of their investments trade in U.S. markets as well. In contrast, the higher volatility, higher trading costs, lesser investor safeguards, and greater sociopolitical risks, along with the impracticality of hedging currency risk, make it prudent for U.S. investors to commit a smaller portion of their capital in emerging markets.

Bringing It All Home

Exhibit 3 tracks the performance of this fully diversified portfolio over the past two decades. The fully diversified portfolio would have lost money only twice—a better record than almost all its constituent asset classes. These two losses—one of them only 0.2%—were also the slightest of the group. The portfolio was saved from greater disaster because bonds made money during bad years for stocks. This optimal blend would have returned at least 10% in 13 out of 20 years (and in two of the other seven returned 8.8%). This portfolio would have compounded at almost 13% a year, growing $100,000 to $1.3 million before taxes, better than all its constituent asset classes except for U.S. stocks, but with far less risk. Perhaps most significant given the current market environment, is that the fully diversified portfolio outperformed U.S. stocks in every major bear market of the past 20 years. In two of the periods, January 1981 to July 1982 and April to December 2000, it actually made money.

	Return	Volatility
Stocks	11.9%	14.8%
Bonds	7.4	5.7
Cash	6.2	0.7