Or so it seems. Even Marilyn Monroe (who starred in the 1953 classic Gentlemen Prefer Blondes) would have approved. So many 401-K and IRA investors are switching to U.S. Treasuries as the preferred investment rather than stocks.
They may be overlooking the volatility of longer maturity Treasury bonds. Many investors are using U.S. Treasurys like a bail bond. They get out of equities and are on good behavior. When they go to “total return court”, however, will they be found neglectful? Or will they jump bail and run back to the stock market? As a recent article in the Wall Street Journal stated, “For the time being, psychology trumps fundamentals. Right now, Treasurys are priced for the same kind of perfection that stock investors banked on in 1999”.

Much like the stock market, however, there really is no “bond market”, just a market of bonds. And there are many types of bonds with differing characteristics and risk levels to choose from. Fixed-income instruments are rated by several agencies as highest in quality (AAA) or very low quality (C). Maturities can range from 90-day notes to perpetual bonds. Sometimes the coupon (the amount that is paid semi-annually) is not fixed and varies with the inflation rate as in the case of the TIPS (Treasury Inflation Protection Securities). In other cases there is no semi-annual coupon at all during the intervening years until the final payoff at the maturity date. This is known as a zero-coupon bond.

Can you make real money with bonds? This is a question we are often asked, and so, with the benefit of hindsight, we’ll show how it can be done. There is a great difference between buying and holding a bond to its maturity, and active bond management. The active manager will weight his portfolio with bonds that possess the characteristics that will optimize returns over a given horizon. If inflation is expected to rise over the near term, the manager will skew the portfolio to the shorter maturities. If corporate profits begin to decline, the manager will emphasize U.S. government debt over corporate debt. Furthermore, if he or she believes that long term interest rates are likely to drop, then they might use a zero-coupon U.S. Treasury bond that is extremely price sensitive to changes in bond yields.

To demonstrate how this might work, we’ll introduce you to the “perfect portfolio”. This fixed-income portfolio holds only the best performing bond type during a calendar quarter. The choices are zero-coupon Treasury bonds (30 years), Treasury bonds (10-15 years), Treasury notes (3-5 years), corporate bonds (3-5 years), Baa-rated corporate bonds (3-5 years), A/Baa-rated corporate bonds (10-15 years) and High Yield bonds (usually 6-8 years and rated BB or below). Chart 1depicts two perfect portfolios, one that excludes the zero coupon bond as a choice and another that includes the zero coupon bond. The performance of these “best bond” portfolios and some of its components can be seen below.

EXHIBIT 1
QUARTERLY BEST BOND PORTFOLIO
1988-2002

(*) A measure of quarterly volatility related risk. Higher deviations equate to higher risk.

Unfortunately, few if any portfolio managers can “pick” the best bond to invest in during any one quarter on a consistent basis. However, since the investment process deals mainly with macro-themes and economics, it may not be as difficult a task as picking the correct equity industry group to invest in during a given quarter. The reason for this is that the most difficult part of the equity equation is the ever-changing coupon (earnings) and the fact that there is no termination or maturity date. For the most part, bonds have a fixed-coupon and a maturity date, making the expected return analysis much simpler.

What if the portfolio manager had to pick the best bond for the year ahead at the beginning of the year rather than for each quarter? Chart 2 shows the comparative performances generated by our two annual “best bond” portfolios and the S&P 500. This is a far less daunting task than making the correct decision at the beginning of each quarter.

EXHIBIT 2
ANNUAL BEST BOND PORTFOLIOS
1988-2002

The key to the differing returns when zero coupon bonds are included has to do with hindsight. In our analysis, it was easy to document when zero’s provided the dominant return during a quarter or a year. Is forecasting a zero’s movement difficult? If you can forecast the change in rates for a 10-15 year Treasury bond (of the coupon variety), you should be able to discern whether or not the more volatile zero-coupon Treasury is worth the risk.

Chart 3 depicts the year-over-year rate of return from both of these bonds. You will note that the swings of the zero-coupon Treasury are simply more magnified than those of the 10-15 year Treasury. Thus there is more volatility risk in “buying and holding” long-term zero coupon bonds when compared to similar maturity coupon-bearing bonds. On the other hand, there is a greater certainty that the stated yield to maturity will be what is actually realized if held to maturity. (In other words, with zeros there is no uncertainty about the rate at which future interest payments will be reinvested.) You may realize significant upside potential if you invest in zero coupon bonds after a period of negative performance.

Few portfolio managers have the luxury of only investing in just one type of bond. However, unlike equity managers who have a broad palette of stocks with many different characteristics, bond managers for the most part are impacted primarily by the changes in interest rates. If we observe how the four-quarter returns between one bond type and another are associated, we see that the levels of correlation (known by the notation “r-squared”, with 1.0 being perfect) are quite high. The exhibit below shows how most of the bond types in our study correlate with the 10-15 year Treasury bonds.

EXHIBIT 3
FOUR QUARTER RETURN CORRELATIONS
1989-2002

(*) adjusted r-squared

It’s notable that high yield bonds (often called junk bonds) are not well correlated with the movements of most other bond types. As a matter of fact, high yield bonds are modestly correlated with returns of the S&P 500 (r-squared=.285). From a fundamental standpoint, an improving economy is of greater benefit to high yield bonds than to other bonds. This is because an improving economy enhances the ability of leveraged corporations to pay down their debt and reduce the risk of default. According to Moody’s Investors Service* , the default rate of speculative-grade debt peaked in January 2002 and is expected to decline from 2002’s 8.3% to 6.9% at the end of 2003. The highest default rates occur in bonds rates “B” and below. Over the last year, telecommunication-related bonds accounted for 54.6% of defaults by volume. This is why bond funds rather than individual bonds are the preferred way to invest in high yield bonds. These funds typically hold 250-300 issues with average maturities of 6-8 years. The average quality rating is what investors should monitor when investing in these funds.

If one is interested in reducing bond portfolio volatility by including a low-correlating asset type, then high yield bonds fit the bill. As with any asset, timing can be the key. Most recently, a well-regarded gentleman by the name of Warren Buffett said he too preferred bonds and was eschewing the stock market. A recent newspaper headline that Buffett was buying “junk bonds” sent many investors scurrying into this sector over the past two weeks. Let’s see what might have interested Mr. Buffett.

Chart 4 shows the relationship between yields on Moody’s Baa corporate bonds and ten-year U.S. Treasury bonds. This ratio oscillated between fairly well defined ranged over the last fifty years. In 1999, as more and more corporate debt was brought to market, the relationship began to deteriorate. Since this is a simple ratio, one might conclude that Treasury bonds are unusually low or Baa bond yields were too high. By March of 2001, one year after the peak in the stock market, Treasury yields began to plummet while Baa yields remained constant as the default rate began to rise much as it had during the 1990 recession This drove the yield ratio up to unprecedented levels. Baa yields remained steady until March 2002, when the default rate began to decline.

Chart 5 depicts the relationship between High Yield and A-Baa corporate bond yields, as well as the 10-15 year Treasury yield. Since 1988, the High Yield/A-Baa yield ratio has traced a sinusoidal pattern. Currently the ratio is at the high end of the range and appears to be headed down as the default rate is expected to continue its decline in 2003. This suggests that High Yield debt should outperform low-end investment grade debt over the next twelve months. Furthermore, the relationship between High Yield debt and Treasury 10-15 year debt is very extreme, but this is probably more a function of Treasury debt being unsustainably low.

SUMMARY

In contrast to how many investors view bonds, the active management of fixed-income portfolios can yield handsome results. Over the past fourteen years, a portfolio of “best bonds” selected at the beginning of a quarter yielded spectacular results. If the “best bond” was picked at the beginning of each year, results well in excess of the S&P 500 were achieved. Interestingly, portfolios that want to increase stability without giving up yield could achieve this through the addition of high yield bonds to their portfolio. Some well-known investors are allocating more money into this sector, and with good reason. High yield bonds are generating income yields well in excess of A-Baa corporate bonds, and significantly above Treasury bonds.

John K. Dolan

Brian E. Schaefer

S&P 500: 801
U.S. Treasury bond (10 Yr): 3.58%
March 12, 2003