Don’t fear a flat yield curve

What, me worry?
What, me worry?
Image: Reuters
We may earn a commission from links on this page.

This economic expansion may be the most robust in recent American history. It has survived shocking election results, the threat of trade wars, the unravelling of the liberal world order, and much else besides. The last recession ended in 2009, almost 10 years ago. Surely something will bring the good times to an end. Could it be the yield curve?

What’s the yield curve?

The yield curve plots interest rates by maturity of government bonds. A “healthy” yield curve slopes up because longer-term bonds should offer a higher interest rate than short-term bonds. Long-term bonds are riskier because they tie up your money for longer and their interest payments may not keep up with inflation. Normally, they pay investors more interest to compensate them for that risk. So, when a 10-year bond pays the same, or less, interest as a one-year bond, people worry something is wrong. An inverted curve has preceded all but one recession since the 1960s.

Market watchers are getting nervous because the yield curve is starting to flatten, as the Fed pushes up short-term interest rates but long-term rates aren’t responding in kind. Fed officials have even started talking about setting rates with the intent of keeping the curve from inverting, a view that could potentially conflict with the central bank’s traditional mandate.

Beware of spurious correlations

An inverted yield curve preceding a recession appears to be a predictable correlation. But no one knows why, exactly. Some speculate that it signals lower expectations about future growth and inflation. But there does not exist a compelling, clear mechanism that can describe why an inverted curve leads to less economic growth.

Typically, an economic relationship, without a reason for being, eventually breaks down. Take the Phillips curve, which predicts that inflation goes up when unemployment goes down. In the 1950s, economists noticed this was normally they case, but did not know why. Then, in the 1970s there was high inflation and unemployment at the same time; today, there is both low unemployment and inflation. In the 1970s, economists could not explain what went wrong with their assumptions because they did not know why the Philips curve worked in the first place. A better theory was developed instead.

The yield curve-recession relationship may be another correlation without a clear story of causation, which means it will probably one day leave economists and policymakers befuddled, too. It may have a good track record, but it does not have a long history. There have only been seven recessions since the early 1960s; it is hard to draw a firm conclusion from so few observations. Also the relationship does not hold up in other countries.

What do we mean by “inverted,” anyway?

Stephen Williamson, an economist at the University of Western Ontario, points out we might be fixated on the wrong points on the yield curve. Recessions are normally preceded by a decline in the difference between two-year bonds and three-month notes. Recently, the flattening has been mostly at the longer end of the curve, beyond 10 years. Williamson argues long-term bonds are not as reliable a predictor of recessions as the shape of the curve at shorter durations. “If you were worried about an oncoming recession right now, based only on yield curve observations, you shouldn’t be,” he concludes.

So why are people worried about the yield curve?

What isn’t in doubt is that the yield curve, normally only interesting to traders and pension nerds, has been getting a lot of attention lately. It could be everyone is nervous about the US economy approaching the end of the cycle. The stock market keeps going up, unemployment is about as low as it can go, and there are few expansions on record as long as the current one. Many people are looking for a sign—any sign—that the cycle is about to turn, and it could be the yield curve. Once again, the curve could signal that a recession is coming, but there’s no certainty that it will come next week, next year, or in another 10 years.