If you are feeling a little “whipped around” lately by roller-coaster interest rates, you are not alone. A year ago, Bloomberg News surveyed 53 economists. The median estimate predicted the 10-year Treasury note would yield over 5 percent by the end of 2004. It finished the year yielding 4.22 percent. That’s a pretty big miss, but it is hardly unusual.
The Wall Street Journal has published a similar survey of economists semiannually since 1982. Bianco Research pointed out in July 2003 that over the life of the survey, the consensus forecast has correctly predicted the direction of interest rates only 30 percent of the time. Most economists are highly trained, very intelligent and have access to thousands of statistics and data series. It is simply a difficult task to accurately predict the future direction of interest rates and even harder to predict the magnitude of change. Last year was certainly no exception.
For most of the year, bond traders focused on the nonfarm payroll employment number as an indicator of economic strength. In February, after posting a three-month average of only 83,000 new jobs, all the talk was of the “jobless recovery.” Expectations were therefore very low when the Bureau of Labor Statistics reported 353,000 new jobs in March. Bond yields rose dramatically in April and May while forecasters ramped up expectations to 200,000-plus new jobs per month. That’s when the June and July numbers came in below 100,000 jobs, and bond yields plummeted back down toward their lows. Many people have been surprised at how wrong the collective wisdom of so many brilliant minds can be.
A recent study conducted by Goldman Sachs sheds some light on the difficulties of forecasting job growth. They conclude that the average nonfarm payroll number over the last eight years was 113,000 new jobs, while the average consensus forecasting error was almost 100,000 jobs. The bottom line is this: Beware of forecasts and be prepared for the unexpected!
How does one prepare for the unexpected? By using flexible investment strategies that react to changing conditions and deliver good results in all market conditions. Most investment strategies are designed to perform well under a given set of assumptions. In fixed income, many of those assumptions revolve on the shape of the yield curve and the direction of rates. To add value, portfolio managers often develop a theory about future interest rates and implement strategies to optimize returns under those assumptions.
However, as we documented earlier, it is very difficult to accurately predict the direction, magnitude and timing of interest rate movements. If the manager guesses incorrectly, serious repercussions may result. For instance, a long duration strategy can produce a 15 percent principal loss in a short amount of time if long-term interest rates rise 1 percent.
Conversely, going to cash will lead to opportunity cost (from less income and foregone capital gains) if interest rates fall. Barbelling (combining short and long maturities) has been popular of late, but it works best only when the yield curve flattens significantly. It typically underperforms other intermediate strategies if the yield curve moves in a parallel or steepening fashion.
Laddering to the rescue
Thornburg has chosen to use laddering because it is an all-weather strategy that has distinct advantages over other bond strategies. Here is a brief discussion of what we believe to be the three principal advantages of laddering.
Yield advantage: A typical Thornburg limited-term bond ladder will consist of bonds that mature from one to 10 years. When the ladder is first assembled, one must buy 10 percent one-year bonds, 10 percent two-year bonds, etc., out to 10 percent 10-year bonds. The average yield of the portfolio will be a blended yield of all the bonds in the ladder—in the example above, 2.9 percent. However, the portfolio yield doesn’t stay there.
If we assume constant interest rates, the portfolio yield will climb every year for nine years. Why? Because the lowest yielding bonds at the bottom of the ladder will mature and be replaced by higher yielding bonds at the top. For instance, after one year, the bottom rung of the ladder, yielding 2.08 percent will mature, and be replaced by a new group of 10-year bonds yielding 3.61 percent. Even though interest rates did not change, the yield on 10 percent of the portfolio just went up by 153 basis points.
If you take this process to its logical conclusion, your laddered portfolio will end up with the yield of a 10-year bond, but a risk profile will have more like a five-year bond.
Price advantage: The front end of the yield curve is almost always the steepest part of the curve. That means that three-year bonds yield significantly less than five-year bonds, which yield less than seven-year bonds, etc.
A Thornburg bond ladder typically involves buying bonds and holding on to them for years. If one holds a bond that was originally purchased as a seven-year bond for two years, it becomes a five-year bond. If interest rates are constant, that bond, which was originally bought to yield 3.22 percent (see example above), will be priced to yield 2.81 percent. If one prices a 3.22 percent bond to yield 2.81 percent, then the bond’s price must go up by 1.9 percent.
In this way, a bond ladder is able to capture price appreciation as bonds in the ladder age and move closer to maturity. This happens even though we assume interest rates don’t move at all. If interest rates rise, the same effect can work to accelerate the price recovery of one’s bond portfolio.
Risk control: Since a laddered portfolio spreads bond maturities over multiple years, only a small portion of your portfolio (10 percent in the example above) will mature over any one period. In other words, you only have to reinvest a fraction of your portfolio at one time.
If interest rates happen to be very low, you can retain the higher yield on most of your portfolio through that period. Conversely, if interest rates are high, you are able to reinvest maturing bonds into the higher yields. The key is that your reinvestment process is controlled, and your portfolio yield should change only gradually over time.
Many people consider bond laddering to be a method of dollar-cost averaging into the bond market. This is because it limits the risk of timing the market poorly when you build and maintain a bond portfolio.
These advantages work to benefit a laddered portfolio whether interest rates go up, down or sideways. Individual investors and investment professionals have a consistently spotty record of timing the market and predicting future interest rates. This is why we believe laddering is the best strategy for risk-averse investors who believe it is important to be prepared for the unexpected.
- The prices of bond funds fluctuate relative to changes in interest rates, with principal values decreasing when interest rates rise and increasing when interest rates fall.
- A Treasury note is a debt obligation of the U.S. government backed by the “full faith and credit” of the government. Notes are intermediate to long-term investments typically issued in maturities of two, five and ten years. Interest is paid semiannually.
- A basis point is a unit for measuring a bond’s yield that is equal to 1/100th of 1 percent of yield. 0.01 percent = 1 basis point.
- A bond whose yield increases from 5 percent to 5.5 percent is said to increase by 50 basis points.
George Strickland is bond portfolio manager and managing director at Thornburg Investment Management in Santa Fe, New Mexico. You may reach him at 800-533-9337.