Can curves predict recessions?

Over a year ago, hopes were high for a post-pandemic surge in activity, encouraging economists and optimistic investors to wager that the U.S. was entering a new Roaring ’20s.

“What a difference a year makes,” says Neil Shearing of Capital Economics. “The narrative has now flipped as concerns about recession have spread,” prompting Shearing to ponder: “Is this latest story any more convincing than last year’s?”

Given that the U.S. economy contracted at a 1.4% annual pace in the first quarter and as the Federal Reserve continues its inflation-fighting rate hike strategy, there have been heightened concerns about a potential recession and how the relationships between bonds are predicting just that outcome.

A yield curve inversion is the unusual market condition when it costs more to borrow in the short term than the longer term.

Typically, it should be cheaper to borrow for shorter periods than longer ones, because lots of things can happen in the future. That’s why bond buyers (lenders) usually demand higher rates to compensate for the additional risk of longer terms. So, in most cases when you buy a 10-year bond, the interest rate is higher than when you buy a two-year one.

But when short-term interest rates are higher than long-term rates, the yield curve inverts — meaning that it slopes downward, which is what has recently occurred.

Tea-leaf readers think that the inversion means that investors are worried the Fed will not be able to thread the needle of increasing rates without throwing the economy into a recession. So, they dump short-dated government bonds and load up on the longer-dated ones. The thinking is that the Fed will raise rates for the next couple of years and then will be forced to do a 180, after the economy slows.

Historically, when the relationship between two- and 10-year government bonds inverts and stays that way for three to six months, it can presage a recession.

Shearing points out that “the curve has inverted ahead of every recession in the U.S. over the past 50 years, with only one false positive (in 1998). It’s therefore about as good a recession indicator as we’re going to get. Ignoring the yield curve means betting against history.”

But inversions may not always be the Magic 8-Ball when it comes to recessions.

For example, the curve inverted in 2019, but it would seem far-fetched to attach the outcome of a once-in-century pandemic-induced recession to that inversion.

What’s more likely is that in 2019, bond investors got a little spooked about the future. Had the two-month COVID recession not occurred, we may have used 2019 as proof that the inversion/recession link was broken.

Additionally, the Fed’s big bond buying campaigns over the past dozen years may be distorting the yield curve, which means the inversion “may have lost some of its predictive power,” Shearing says.

But this inversion perversion underscores just how spooked we are about recessions in general. While nobody wants to see a return to high unemployment and human suffering, recessions are natural occurrences of the economic cycle. Sometimes the contraction and subsequent recovery last a long time (the Great Recession) and sometimes the damage is deep, but the length is short (the COVID recession).

So far, there are signs of a slowdown, as high inflation and rising rates eat into corporate profits and personal spending, putting inversion-adherents and Fed watchers on high alert.

“The challenge will be for the Fed to cool domestic demand without sending too much of a chill through the labor market,” says Grant Thornton chief economist Diane Swonk. “Getting policy ‘just right’ is no easy feat. Goldilocks only exists in fairy tales.”