Recent cash flows into bond mutual funds and exchange-traded funds have been very strong. A slightly higher percentage of bond cash flows has gone to short-term funds. This is fairly atypical for an environment of extremely low short-term yields.
One of the likely catalysts for this trend has been increased demand from money market investors in search of higher yields. With the Federal Reserve maintaining its federal funds rate target close to 0%, monetary policymakers have made it extremely difficult for many savers to generate sufficient income from their money market accounts. In this sense, savers unfortunately remain the “sacrificial lambs” of U.S. monetary policy as the Federal Reserve attempts to stimulate other segments of the economy.
Another probable influence is increasing concern among bond investors that mounting government debt levels will eventually drive up longer-term U.S. interest rates, which at present are below their historical averages (see Figure 1). In addition, the futures market expects the Federal Reserve to begin raising short-term rates before the end of 2010 as the U.S. recovery strengthens.
Viewing these concerns together, some bond investors may hope that the total returns on shorter-duration funds will be relatively insulated in the event that both short- and long-term rates rise by the same amount (that is, a parallel upward shift in interest rates). In light of these uncertainties, it’s natural for bond investors to wonder whether they should act defensively by reshaping their fixed-income allocation with a narrow or “surgical” focus on mitigating risk. To provide better perspective and grounds for discussion, we begin by examining how the market expects interest rates to move and how various government bond indexes might perform if those expectations are met.
Figure 2 shows the bond market’s expectations for future interest rates along the entire yield curve. These implicit expectations—often referred to as the forward yield curve, or simply forward rates—can be derived from current U.S. Treasury bond prices. In essence, the forward curve can be considered the set of “break-even” yields that equalize the rates of return from Treasury bonds across the entire maturity spectrum. An important (and often misunderstood) implication of Figure 2 is that if the yield changes of the forward curve are realized, then all Treasury bonds—regardless of their maturity—will earn the same holding-period return.
Given the dramatic steepness of today’s yield curve (as measured by the difference between the 10-year Treasury yield and the yield on the three-month T-bill), the bond market expects the yield curve to flatten significantly in the years ahead. The “bear flattening” scenario in Figure 2 shows that the majority of the interest rate adjustment is expected to occur in short rates, not long rates. As an example, the two-year Treasury yield is expected to rise strongly over the next five years as the Federal Reserve normalizes monetary policy, driving the two-year note from its February 2010 spot yield of 0.81% to 5.28% in February 2015. The yield on a 20-year constant-maturity Treasury bond is expected to rise less dramatically, from the February 2010 spot yield of 4.43% to 5.56% in February 2015.
In short, the Treasury security market has already “priced in” a Fed tightening cycle that flattens the yield curve in such a manner that the expected return on, say, a short-term Treasury portfolio would be approximately the same as for a longer-duration Treasury portfolio. As illustrated in Figure 3, this flattening of the yield curve would be similar to what occurred following the end of the last low-rate environment (2003–2004). Starting in 2004, the Federal Reserve raised short-term rates in stages from 1.00% to 5.25%, while the yield on the 10-year Treasury hardly changed. The Fed chief then, Alan Greenspan, called this situation a “conundrum” because it differed from other Fed rate-tightening cycles, such as that of 1994, when longer-term yields rose almost in tandem with the rise in the federal funds rate. However, others have argued that the conundrum episode was precisely how a tightening cycle should operate under a credible central bank that effectively anchors long-term inflation expectations—the primary driver of long-term bond yields.
A natural question, of course, is whether the bond market’s expectations for future long-term Treasury bond yields are reasonable. In light of the concerns over U.S. fiscal deficits and the Fed’s exit strategy, many investors may find it perplexing that a 10-year Treasury bond yields less than 4% today and is expected to yield only about 5.6% in 2015 (recall Figure 1). More pointedly, don’t long-term rates have to rise more dramatically than that in response to the large and growing national debt? In short, the answer is: No, not necessarily.
It is important to recognize that the relationship between U.S. government debt levels and long-term government bond yields is mixed and has varied dramatically over time. The average correlation between long-term bond yields and federal debt levels has been zero.
In fact, were one to graph the current relationship between debt-to-GDP ratios and long-term government bond yields across countries, it would show a strikingly similar correlation of zero. The reason for the general absence of a close association between debt and interest rates is that there is only a weak link between deficits and inflation, at least in developed markets. The weakness of the link is attributed in part to cross-country differences in central-bank credibility, economic size, domestic private savings rates, and perception of future fiscal prudence.
Japan, for instance, has one of the lowest long-term bond yields in the world despite having the highest debt-to-GDP ratio. An explanation is that Japan’s long-term inflation expectations remain close to 0% (if not outright deflationary), and investors (mostly Japanese private citizens) have been willing to fund the government debt.
A significant finding is that the expected upward pressure from the fiscal deficit on long-term bond rates in 2009 has been offset (at least thus far) by increased bond demand arising from a higher domestic savings rate. By our estimates, this recent suppressant of long-term bond yields has been as powerful as the so-called “global savings glut” that Fed Chairman Ben Bernanke and others believe helped to keep long-term yields low throughout the past decade. According to the Federal Reserve’s Flow of Funds data, households and U.S. mutual funds owned approximately 20% of all U.S. Treasury securities outstanding in the third quarter of 2009, versus 14% in 2007 before the crisis.
If the U.S. savings rate stays elevated in the years ahead, future long-term bond yields may not rise as strongly as some investors now fear. The behavior of Japanese interest rates today and of U.S. rates during World War II are reminders of the powerful influence that a higher domestic savings rate can have on a government’s borrowing costs. Another critical factor going forward will be the effectiveness of future monetary and fiscal policies in maintaining stable long-term inflation expectations. This is because inflation expectations are the largest single component of long-term interest rates.
We decompressed interest rates to shed light on factors influencing the current levels of interest rates and to provide a basis for assessing the future economic scenario that is priced into today’s bond valuations (and hence, into the forward yield curve). By combining our interest rate decomposition calculations with the market’s expectation that the 10-year Treasury yield will rise to approximately 5.6% in 2015 (recall Figure 1), it can be shown empirically that the anticipated bear flattening is consistent with the following scenario over the next five years:
As such, the bond market’s expectations for the future shape of the yield curve seem reasonable. Of course, history suggests that rates will likely evolve differently than what is expected today. Indeed, the Treasury’s forward yield curve—as with other interest-rate forecasts—has been a poor predictor of actual future rates. This has often been the case for both short-term rates and long-term rates.
In light of this information, how should bond investors think about the risks to current market rate expectations?
Recognizing that the number of distinct future yield-curve scenarios is nearly infinite, we chose five potential scenarios from 10,000 simulations generated by the Vanguard Capital Markets Model for a calculation of implied returns. Table 1 presents the results.
Most broadly, the scenarios that produce the highest short-run returns would be expected to produce the lowest long-run returns. A good example is the so-called double-dip scenario; for long-term investors, Scenario 4 would likely be the most troubling. In this scenario, the present low-rate environment persists indefinitely as the economy falls back into recession and suffers from a decade-long malaise similar to what Japan has experienced for the past two decades. Indeed, it is this very outcome that the Federal Reserve has endeavored to avoid.
Conversely, scenarios in which rates rise more in 2011 than is currently expected actually produce the highest relative nominal 10-year returns. For instance, short-term bond indexes have lower returns in Scenario 3 than in Scenario 1 because the Fed raises rates more quickly and aggressively than is presently expected by the market. Yet in Scenario 3, long-run inflation expectations remain well anchored, and thus intermediate-term bond indexes have similar 10-year returns in both scenarios.
Indeed, the scenario that is perhaps the most feared by many bond investors (Scenario 5) is also the one with the highest expected return over a 10-year horizon. Naturally, under this “fiscal crisis” scenario all bond indexes would be expected to suffer significantly negative returns in the short run as interest rates rise sharply. However, over time, the higher Treasury and corporate yields would provide a higher absolute income stream—as many fixed-income investments eventually did in the 1970s and early 1980s.
A key lesson of the global financial crisis is that implementing a too-narrow or surgical bond allocation (such as by shortening duration or investing solely in riskier bond instruments) involves important trade-offs that may expose bond investors to unintended yield-curve or market risks while potentially depriving them of higher future income streams. These trade-offs are clearly evident in the range of potential interest rate scenarios that we have depicted in Table 1.
|Scenario 1: Treasury forward yield curve is realized.|
|1–5 Yr. Treasury Index||-0.6||1.2||3.1|
|5–10 Yr. Treasury Index||-1.8||1.2||3.2|
|1–5 Yr. Corporate Index||-0.5||1.4||4.0|
|5–10 Yr. Corporate Index||0.9||2.3||4.0|
|Scenario 2: Federal Reserve on hold for longer than expected.|
|1–5 Yr Treasury Index||0.7||1.0||3.1|
|5–10 Yr. Treasury Index||2.8||1.6||3.3|
|1–5 Yr. Corporate Index||3.1||1.0||4.2|
|5–10 Yr. Corporate Index||6.2||3.6||5.5|
|Scenario 3: “Preemptive” Federal Reserve is more aggressive than expected.|
|1–5 Yr. Treasury Index||-1.1||3.4||3.7|
|5–10 Yr. Treasury Index||-4.2||1.8||3.2|
|1–5 Yr. Corporate Index||-2.3||3.4||4.0|
|5–10 Yr. Corporate Index||0.8||3.5||5.4|
|Scenario 4: Double-dip scenario in 2011; Japan-type economic stagnation thereafter.|
|1–5 Yr. Treasury Index||1.4||2.0||2.1|
|5–10 Yr. Treasury Index||11.9||4.0||2.5|
|1–5 Yr. Corporate Index||0.5||2.4||3.2|
|5–10 Yr. Corporate Index||8.5||6.4||5.1|
|Scenario 5: Fiscal and inflation concerns accelerate dramatically; all rates rise.|
|1–5 Yr. Treasury Index||-1.4||2.6||3.9|
|5–10 Yr. Treasury Index||-12.1||0.8||4.5|
|1–5 Yr. Corporate Index||-2.9||3.0||4.7|
|5–10 Yr. Corporate Index||-12.4||1.8||6.5|
|*Implied future annualized returns for one year ending 2/2011, five years ending 2/2015, and 10 years ending 2/2020.|
|Sources: Barclays Treasury and Credit Index benchmarks, Barclays Capital; Bloomberg; Federal Reserve; and Vanguard.|
These varied—but certainly possible—rate scenarios attest to the high degree of uncertainty surrounding the future direction of economic growth, the deficit, inflation, and interest rates. Indeed, the difficulty of correctly forecasting not only which (if any) of these scenarios will unfold, but also precisely when, is a powerful reminder that focusing on interest rate moves and short-term changes in bond prices can be counterproductive. To us, the range of potential outcomes in Table 1 would seem to support greater fixed-income diversification in the years ahead, not less.
The performance of various segments of the bond market over the past several years underscores the benefits of a broadly diversified fixed-income portfolio regardless of the future direction of interest rates.
Over the long term, interest income—and its reinvestment—accounts for the largest portion of total returns for many bond funds. The impact of price fluctuations can be more than offset by staying invested and reinvesting income, even if the future is similar to the rising-rate environment of the late 1970s and early 1980s. According to Federal Reserve data, the yield on the 10-year Treasury bond more than doubled over this period, rising from approximately 6.9% in December 1976 to as high as 15.3% in September 1981. Yet the hypothetical $1 million investment made in 1976 would have grown to more than $2.0 million by the end of 1983—not necessarily a disastrous outcome given the period’s secular rise in interest rates. Moreover, the higher level of interest rates in the early 1980s subsequently fell as inflation expectations declined, setting the stage for even higher bond returns over the following decade.
Roger Aliaga-Diaz, Ph.D., Don Bennyhoff, CFA, Andrew J. Patterson, and Yan Zilbering also contributed to this article.