- Predictions of negative returns from bonds once the Federal Reserve (Fed) starts pushing short-term rates upward do not give sufficient weight to the role that today’s exceptionally steep yield curves could play in future bond market performance.
- During six postwar periods of Fed tightening, yield curves began steep and flattened significantly at the end of the tightening cycle.
- In my view, today’s steep yield curves may help mitigate the negative effects of rising rates.
It has become commonplace to hear predictions of negative returns from bonds once the Fed starts pushing short-term rates upward. I would argue that those predictions do not give sufficient weight to the fact that yield curves today are exceptionally steep. While the implications of these curves for future bond market performance are not widely discussed, the history of previous interest-rate cycles would suggest that, largely because of the extreme steepness, the next cyclical rise in bond yields might not be as severe as many observers are predicting.
Six periods of Fed tightening began with steep yield curves that flattened at the end of the tightening cycle
Six cycles in the postwar period in which the Fed began tightening show yield curves that were steep. The first occurred in 1959 and the last in 2004. Over those six episodes, the average spread between yields on 30-year and 2-year Treasuries at the onset of Fed tightening was 170 basis points (bps; 100 bps equals 1.00%). The narrowest spread was 100 bps in 1973, and the widest spread was 275 bps in 2004. That spread is now 335 bps, almost double the average and 60 bps wider than in 2004.
In each of these six cycles, the Treasury yield curve was significantly flatter at the end of the Fed tightening cycle than it was at the beginning. In fact, in most of those cycles, the curve was inverted by the time the Fed stopped tightening. The average 30-year to 2-year slope at the cyclical peak in the federal funds rate (that is, the rate that banks charge each other for the use of federal funds) was -45 bps. The range was from -150 bps in 1979 to 50 bps in 1995. Because of this flattening pattern, the cyclical increases in note and bond yields were much smaller than the increases in short-term rates.
Because the increases in the federal funds rate were quite large, and because a rising federal funds rate will typically pull up yields on other short-maturity securities, the cyclical increases in 2-year Treasury yields were substantial. On average, those yields almost doubled. The increases ranged from 150% in 1959 and 2004–2006 to 60% in 1973–1974. For 10-year notes, however, the average yield increase was only 30%. The range was from 45% in 1959 to only 2% in 2004–2006. For the 30-year bond, the average cyclical increase was 20% and the range was from 35% in 1977–1979 to 2% in 2004–2006. In other words, the yield increases for the 10-year and longer maturities were on average less than one-third the increases in the 2-year yields, and the increases in the 2-year yields were much smaller than the increases in the federal funds rate.
Every cycle is different, so generalizing from past experience is always tricky. Nevertheless, the flattening process has been so consistent when curves were steep that it would be unwise to ignore the potential for significant flattening in this cycle. There was even one instance, in 1959, when the federal funds rate was almost as low at the onset of Fed tightening as it is now. Back then, the federal funds rate was around 0.35%, but the 2-year yield was around 2%, much higher than it is now. The 10-year and 30-year yields were, however, around 3.25% and 3.50%, respectively—much closer to today’s yields. In that tightening episode, the federal funds rate reached 4% and the 2-year yield reached 5%, but the 10-year and 30-year yields increased far less—to around 4.50% for each maturity.
Steep yield curves, in my view, may help mitigate the negative effects of rising short-term rates
A pronounced flattening of the yield curve, along the lines of the average movements over those previous six cycles, would, in my opinion, limit the cyclical rise in bond yields to a fraction of the rise in short-term rates. Moreover, if the rise in the federal funds rate is gradual, as Chairman Bernanke has suggested is most likely, yield curves could flatten as dramatically as they did from 2004 to 2006 (see chart). At that time, the federal funds rate increased from 1.00% to 5.25%, but in predictable 25-bp increments, and bond yields were essentially unchanged over the cycle. As the table shows, despite a substantial increase in short-term rates, total returns from bonds were significantly better than returns from defensive assets such as Treasury bills.
The consensus forecast from members of the Federal Open Market Committee is that even a gradual rise in the federal funds rate might not begin until 12 to 24 months from now. In that event, the incremental interest income provided by unusually steep yield curves would likely be the primary component of total return. Intermediate- and longer-term portfolios would be expected to outperform short-term portfolios. And even as short-term rates rise, the compounding of that interest income, together with the flattening of yield curves, would likely produce better returns from bonds than the most dire commentaries are now predicting. In my view, steep yield curves are an investor’s friend because they may help mitigate the otherwise negative effects of rising short-term rates.
|Cumulative total returns (%) 6-1-04 to 7-1-06|
|Treasury 30-year bonds||10.6|
|Corporate bonds, 15+ years||10.1|
|Municipal bonds, 10+ years||14.6|
Source: Merrill Lynch, Bloomberg. Securities issued by the U.S. government are guaranteed as to the timely payment of principal and interest.
Past performance is no guarantee of future results.
U.S. Treasury yield curves, 6-1-04 to 6-30-06
Past performance is no guarantee of future results.