What The Yield Curve Is Telling UsKaren Pennar
Rarely do economists insist they have a foolproof method for predicting future economic activity, but these days more and more prognosticators are putting their faith in the signals being sent by the yield curve--a simple, connect-the-dots rendering of where interest rates are along the maturity spectrum from 3-month Treasury bills to 30-year Treasury bonds. Trouble is, much like a Rorschach test, different people looking at the same curve come up with different economic forecasts. Some believe the yield curve is saying growth will slow down in the second half. Others believe it's predicting a sharp slowdown or even recession by early 1996. And an academic who did extensive yield-curve analysis in the 1980s says the curve is telling him that growth will be fairly strong in 1995.
Disagreement comes over the significance of one incontrovertible fact--that the yield curve has been flattening lately. That means short- and medium-term interest rates have been moving higher, while long-term rates have been steady. Historically, such a flattening of the curve has predicted slower growth, while an inversion of rates--when short-term rates move above long-term rates--has signaled a recession to come.
HISTORY LESSON. In mid-December, the rate on the 5-year Treasury note climbed above the rate on 10-year bonds, and it has remained at or close to the rate on 10-year instruments since then. And while the 3-month and 6-month rates remain well below long-term levels, they nonetheless have been rising thanks to market pressure and action by the Federal Reserve Board. Moreover, the spread between the 1-year rate and the 10-year rate has been narrowing lately. It may be difficult for anyone but yield-curve aficionados to recollect such shifting relationships between rates and what they presage, but one case may jog people's memories: In 1980, short rates shot way past long rates and touched 20% for a time. A deep recession ensued.
Technically, only one segment of the yield curve--the 5- to 10-year stretch--inverted last month. But that's enough for David H. Resler, economist at Nomura Securities International Inc. That inversion tells him that "we're headed for significantly slower economic growth in the second half" on the order of about 11/2% or so, vs. 4% for the second half of 1994. Historically, he says, an inversion in the middle segment of the yield curve has always signaled slower growth about six months down the road. And, he adds, it generally signals a recession or growth recession--very slow growth after a period of robust growth--18 months after that inversion.
The theory behind yield-curve analysis is a simple one. Interest rates reflect the price of money, and in so doing, they convey information about people's expectations about growth and inflation and their preferences about holding money. Thus when long rates are rising faster than short rates and the yield curve has a positive slope, people are expecting strong growth and higher inflation and are demanding higher interest rates on longer-dated securities. When short rates rise faster than long rates and the curve flattens or inverts, that means they expect weaker growth or recession and lower inflation, so they accept lower rates on long-term bonds.
Thus it's logical, says Campbell R. Harvey, a professor at Duke University's Fuqua School of Business, that inverted yield curves are clear "omens" of recession. Harvey, in his doctoral dissertation, studied the relationship between 3-month and 10-year Treasury rates over several decades. He found that since 1955, an inversion of these rates produced, with a 5-quarter lag, a recession whose duration generally matched the length of the inversion.
There was only one false signal in the mid-1960s. But the Fed was manipulating the yield curve to support the dollar then, and even so, the inversion did precede what was widely deemed to be a "growth recession." Mark Zandi, economist at Regional Financial Associates Inc., says that without heavy government spending at the time, there likely would have been a recession in 1967, as the inversion in 1966 predicted.
TILT? What happens next depends on whether or not today's flattened yield curve tilts over and truly inverts across the maturity spectrum. Katherine R. Hensel, chief investment strategist at Lehman Brothers Inc., looked at the relationship between 1-year and 10-year Treasuries and found that the curve was similarly flat in the early 1960s and in 1971, 1975, 1977, 1984, and 1986. Not all these periods were followed by a yield-curve inversion and recession.
Other analysts also are expecting slow growth, but not recession. RFA's Zandi figures the shape of the yield curve points toward 2% growth in the second half of 1995. But Harvey is counting on only a slight moderation in the growth rate, to about 31/2% to 33/4% by yearend. That's because he doesn't think short rates have moved high enough to slow growth that much. And he isn't looking for a true inversion of short rates vs. long rates anytime soon.
The yield-curve watchers may not always agree on exactly what the flattening of the yield curve means, but all of them, including Harvey, agree on one thing--if short rates rise above long rates, recession is sure to follow. No Rorschach test will be necessary to divine what kind of damage that will do.