More on How Yields Are Throwing a Curve at Investor Expectations

Last week's discussion of basic yield-curve dynamics left many of you begging for more, if my email box is any indication. So this week, let's look at the history of the Treasury yield curve. It helps explain why long-term yields have been rising while short-term yields have been steady.

The Treasury yield curve (please go back and read last week's column if you don't understand what it is) continues to undergo big changes. These moderations reflect changing opinions about inflation, monetary policy and presidential politics.

To briefly recap, the yield of the longest maturity issue, the 30-year bond , is higher than the yields of the 10- and five-year Treasury notes for the first time since January.

As a result, the yield curve, which was completely inverted from January till last week, is now partially reversing that trend. I say "partially" because the two-year note's yield remains higher than the five-year note's, which is a smidge higher than the 10-year note's. But the 30-year bond's yield is no longer lowest of all.

Last week, I explained how it is that a yield curve ever inverts. Why, after all, would investors buy long-term bonds at a yield lower than they could get from short-term bonds? Because they believe that, over time, they will get better total return -- income plus price appreciation -- from the long-term issue than the short-term one. Traditionally, this expectation is rooted in a belief that economic growth will slow, or even turn into recession. (Again, all this is explained more fully in last week's column.)

This year's inversion was different. It happened, in large measure, because of the prospect of a shrinking supply of long-term Treasuries. The federal government started using surplus funds to pay down debt by buying back long-term Treasuries from investors.

Glossed over in last week's discussion (and wondered about in reader emails) was this fact: The way the yield curve inverted this year was fundamentally different from the way the yield curve has inverted historically. And that helps explain why investors are selling long-term Treasuries, even as they begin to entertain the idea that the Fed will lower the fed funds rate sometime next year.

Stick with me here. Several of you asked why investors would sell long-term Treasuries if they expect economic growth to slow, possibly so much that the Fed will cut interest rates. After all, when the Fed cuts interest rates, the bond market rallies, right? And when the bond market rallies, you want to be holding the longest-maturity issues, because they're going to give you the best total return.

As we discussed last week, oil is a big part of the story. Bond investors may expect that growth will slow as a result of high energy prices. But if they also expect that the inflation rate is going higher before it goes lower, they will demand higher yields on long-term bonds, even if they are certain the Fed won't raise rates.

But there's another aspect to the story: For much of bond-market history, an inverted Treasury yield curve has marked the bottom of the market. In other words, when the Treasury yield curve reached its point of maximum inversion (when long-term yields were as low as they were going to go, relative to short-term yields), interest rates were peaking. Conversely, at the point of maximum steepness, yields were troughing.

The reason is simple. The level of short-term yields depends on monetary policy. They rise when the Fed is hiking the fed funds rate and fall when the Fed is cutting it. High short-term yields lead to slow economic growth, and slowing economic growth means declining bond yields. Conversely, low short-term interest rates cause growth to accelerate, and investors to demand higher bond yields.

The investment implications are clear. When long-term yields are much lower than short-term yields, the best bet is to buy long-term issues. And when long-term yields are much higher than short-term yields, the best bet is to buy short-term issues. It's counterintuitive, but for years, it made money.

The problem is, it hasn't worked quite as well lately. That's because the relationship between the slope of the yield curve and the direction of interest rates has broken down to a large extent. Consider this chart.

Yields and the Curve

The thick blue line measures the Treasury yield curve. That is, it measures the difference in yield between the 30-year bond and the two-year note. When it is below zero, the yield curve is inverted (the two-year note's yield is greater than the 30-year bond's). When it is above zero, the yield curve has a normal, positive slope. Where the line peaks, the curve reached its point of maximum steepness, and when it troughs the curve reached its point of maximum inversion (or flatness, in cases where it didn't actually invert).

The red and light blue lines are the yields of the 30-year bond and the two-year note, respectively.

As you can see, from 1977 till the early 1990s, the yield curve reached its point of maximum inversion when interest rates were peaking, and reached its point of maximum steepness when rates were bottoming.

However, since the early 90s, there has been no clear trend, as this next chart (which merely zooms in on the 1994-to-present period) shows.

Yields and the Curve Lately

Of course, since the early 1990s, neither yields nor the yield curve have done much of anything, compared with the fireworks of the previous 20 years. But neither has there been much correspondence between the direction of interest rates and the shape of the yield curve.

Most notably, while the yield curve gradually flattened last year as the Fed hiked rates and yields climbed, it didn't invert until the market started to rally in January.

The yield curve can invert because short-term yields are rising more than long-term yields (as was common for years), or because long-term yields are falling more quickly than short-term yields (as happened in January). Likewise, the curve can steepen either because short-term yields are falling more than long-term yields (the historical norm), or because long-term yields are rising more quickly than short-term yields (as has been happening lately).

The correspondence between interest rates and the yield curve has broken down for two main reasons, says Jim Bianco, president of Bianco Research in Barrington, Ill.

First, because the Fed, which prior to 1994 did not announce changes in monetary policy (instead leaving the markets to figure it out for themselves), now not only announces changes, but explains them.

Pre-1994, Bianco points out, interest rates didn't peak till the Fed finished hiking rates, and they didn't trough till the Fed was through cutting rates. Now that it is easier to predict the Fed's monetary policy moves, the attitude about interest rates is: Why wait? "As the Fed has been more transparent, it's easier to front-run what they're going to do," he says. "Now the market turns halfway through a Fed campaign."

Look at these charts. The first shows the high degree of correspondence between Treasury yields and the fed funds rate prior to 1994. The second zooms in on the 1994-to-present period, where you can see the front-running Bianco is talking about: Yields now peak and trough before the fed funds rate.

Yields and the Fed Funds Rate

Yields and the Fed Funds Rate Lately

The other reason why the correspondence between interest rates and the yield-curve slope has broken down, Bianco says, is because banks, which set interest rates relative to the fed funds rate, play a much smaller role in the credit markets than they used to, while the bond market plays a much larger role. For example, corporations are much likelier now to issue bonds than to borrow from banks.

The level of the fed funds rate still has an important psychological effect on the markets, Bianco says, but its practical effect on the economy is diminished. So it stands to reason that its correspondence to the direction of long-term Treasury yields should be diminished as well.

As a corollary to this, Bianco points out that the corporate-bond yield curve -- the difference in yield between a long-term and a short-term corporate-bond proxy -- has done at least as good a job as the Treasury yield curve at predicting economic activity. In other words, an inverted corporate curve has been a reliable harbinger of a slowing economy, and of a peak in interest rates. As this chart shows, the corporate yield curve is not inverted (long-term yields are higher than short-term ones), and it hasn't been since 1989.

A Tale of Two Yield Curves

Source:Bianco Research

People have been selling long-term Treasuries because they are worried that higher oil prices could lead to higher inflation, and because they now question whether the supply of long-term Treasuries will continue to shrink, leading to further price gains. But also, they are taking profits on a development that, from a historical perspective, is unusual: a rally in long-term Treasuries in the midst of a series of interest-rate hikes by the Fed.

---

Send your questions and comments to [e-mail], and please include your full name.