Bond investors have reaped the benefits of attractive supply-and-demand dynamics, narrowing spreads, and low default rates since the credit crunch of 2008. Whether the decades-long bull market will soon come to an end is up for debate, but what isn't is that rising interest rates can wreak havoc on a bond portfolio's total return.
The inverse relationship between bond prices and interest rates is a central tenet of bond mathematics. As rates rise, new, higher-coupon bonds become more attractive than previously issued lower-coupon bonds. In order to get investors to buy those lower-coupon bonds, prices must fall. Eventually, higher bond yields will be good for investors, but the short-term pain can take investors by surprise.
Experienced investors may look at a fund's duration (a measure of price sensitivity due to changes in interest rates) to determine whether a portfolio is "protected" from rising rates. Duration, however, has a number of shortcomings. It's a good measure of interest-rate risk for non-callable Treasury bonds but becomes less predictive for funds with more credit risk, such as junk bonds, or unpredictable cash flows, such as mortgage-backed securities. In addition, portfolio duration measures assume a parallel shift in the yield curve. Simply, they assume that all rates along the entire yield curve (from one month to 30 years) move up or down by the exact same amount. This is almost never the case in real life.
The graph below shows Treasury yield curves as of Dec. 31, 2013, and June 30, 2014. Since the start of 2014, the yield curve flattened somewhat as longer-term rates fell further than short-term rates. The changes were far from parallel. In fact, one-month and two- and three-year yields actually rose slightly over the seven months. Changing interest rates can cause the curve to flatten and steepen on a daily basis, which affects the prices of varying maturity bonds differently.
- Source: Morningstar Analysts
Though two fund managers may target the same duration, the underlying portfolios may vary widely in their makeup, in turn causing performance to deviate as rates change. Portfolio duration is a complex topic, but a simple example can illustrate this concept.
Assume three zero-coupon Treasury bonds (Bond 1, Bond 2, and Bond 3) held in either Fund A or Fund B, as of Dec. 31, 2013, are combined to create a duration of five years for both funds. Fund A owns the five-year Bond 2 and Fund B owns a combination of the two-year Bond 1 and the 10-year Bond 3 (62.5% and 37.5%, respectively). Using the yield curve rates from Dec. 31, 2013, the tables below show relevant bond data. Because we are assuming a zero-coupon bond, the duration of each bond is equal to its maturity.
Now assume an instantaneous shock to the yield curve that causes the two-year rate to rise to 0.60%, the five-year rate to fall to 1.60%, and the 10-year to drop to 2.5%. The next table shows the effect on each bond's price.
Assuming no changes to the holdings of the funds, the third table shows the change in portfolio value based on the above yield curve shock. Despite the funds starting the year with the same duration (five years), the composition of the portfolio and the nonparallel shift in the yield curve meant that Fund B outperformed Fund A significantly. A drop in the 10-year yield pushed that bond's price higher, only partly offset by the rise in two-year yields, which caused that bond's price to fall. Investors expecting the funds to perform similarly would have been quite surprised by the results.
Of course, this example is oversimplified. Duration is a complex topic, and gaining insight into the composition of a fund's duration is challenging for investors. What's more, the yield curve doesn't always follow the same path when policy rates are on the rise. For example, when the Federal Reserve hiked rates between June 2004 and July 2006, the yield curve flattened dramatically. The one-month Treasury rate started the period at 0.97%, rising to 5.02%, while the 20-year Treasury rate dropped to 5.17% from 5.45% over the two years. But during the rate hike from July 1999 to July 2000, the curve inverted. One-month rates started the period at 4.68% and ended at 6.27%, while 30-year rates started at 6.02% and dropped to 5.79% over the year.
How the yield curve will shift as the Fed begins this round of rate hikes is something of a wild card. But it's important to have a handle on what duration can and cannot reveal about a fund's potential behavior. What's more, understanding that the changing shape of the yield curve can affect a fund's value in a meaningful way can allow investors to create more-realistic expectations around returns.