The signal and the noise.

Photographer: Mario Tama/Getty Images

The Yield Curve Isn't Infallible

Stephen Mihm, an associate professor of history at the University of Georgia, is a contributor to the Bloomberg View. Follow him on Twitter at @smihm.
Read More.
a | A

In recent months, the yield curve has started to flatten, prompting some of Wall Street’s seers to declare that a recession is in the offing while others say this is just a blip. What is this much-watched method of economic divination trying to tell us?

The yield curve is used to assess the future of interest rates by plotting the return paid by bonds of the same quality across a range of maturity dates. It typically slopes upward: Short-term bonds pay lower rates than bonds that mature farther into the future. But this changes when economic clouds appear on the horizon and traders anticipate lower growth and lower interest rates.

The most recent moves in the yield curve have produced a line that is less a curve than a very modest incline, the flattest it has been since 2007. And this isn’t restricted to the U.S.: The yield curve in the U.K. is now the flattest it has been since 2008. These years, of course, marked the beginning of the financial crisis and a nasty recession, and more than a few people believe the flattening heralds another big downturn. But skeptics view the dip as nothing more than a temporary response to the unorthodox measures undertaken by central banks in response to the crisis.

QuickTake Negative Interest Rates

Who is right? It is worth remembering that the predictive power of the yield curve has been quite variable over time and it has been a better gauge under certain conditions than others. That insight can help investors assess whether they should trust the curve at this critical moment.

The idea that the yield curve could predict the future has its roots in the 1960s, when Ruben Kessel of the National Bureau of Economic Research first observed that the difference between short and long-term rates tended to narrow in advance of a recession. In the 1970s, Marty Leibowitz at Salomon Brothers began tracking bond yields on a curve rather than treating each class of bonds separately.

The new methods, according to the financial historian Caitlin Zaloom, “connected what had been previously considered separate packages of time and traded as separate markets. These practices brought life to a new market vision that animated the curve with the health or weakness of the economy to come.” 

In time, academic researchers elevated the yield curve to a leading indicator of the nation’s economic health. It soon became holy writ that a high slope indicated fast growth; a flattening of the curve meant slow growth or even a recession; and an inversion of the curve meant nothing less than a recession.

But the yield curve is more complicated than this simple summary would suggest. Its ability to predict the future is very much dependent on historical conditions.

The economic historian Michael Bordo, in several papers published in the past 10 years, sought to determine when the yield curve functioned better as a predictive tool. To answer this question, he and his collaborator, Joseph Haubrich of the Federal Reserve Bank of Cleveland, set out to construct yield curve data going back to 1875. This was no easy feat: The short-term three-month Treasury bills that are an integral data point for today’s yield curves weren’t introduced until 1929. 

Bordo and Haubrich got around this problem by following the lead of other historians of the yield curve: They constructed a surrogate by plotting the interest rates of high-grade, short-term commercial paper for the start of the curve and high-grade corporate bonds for the other end. Then they looked at whether a narrowing in the spread between these classes of debt served as an accurate harbinger of slowdowns and recessions between 1875 and 1997.

The results are intriguing. They surveyed many distinct periods in U.S. financial and monetary history, from the era of the classical gold standard to the creation of the Fed to the breakdown of the gold standard, its resurrection on the international level under Bretton Woods, and its eventual collapse in 1971.

They found that the utility of the yield curve had a great deal to do with the monetary regimes of the day. In particular, they found that the yield curve worked better as a predictive tool when the monetary system was out of kilter and inflation a problem, and worse when the monetary system was predictable and inflation stable.

In other words, it would seem that the utility of the yield curve is inversely related to the credibility of the monetary regime. When the market thinks that inflation could spiral out of control because of the incompetence of policy makers, the yield curve works pretty well. But when inflation is kept in check, you can’t take the yield curve as seriously.

Which brings us to today. Although we still don’t know for certain what the yield curve is telling us, the possible answers come into sharper focus. If the market is sensing that the Fed doesn’t have a hold on inflation, then perhaps the yield curve is unequivocally signaling a recession. But if inflation isn’t a threat, then the yield curve could be giving us a misleading reading.

Right now, there’s almost no sign that inflation is about to spin out of control. And that, perhaps, is a sign that the crystal ball known as the yield curve is too cloudy to trust.

This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.

To contact the author of this story:
Stephen Mihm at

To contact the editor responsible for this story:
Max Berley at