Beware of “foolproof” recession indicators. For years, people have been telling me that an inversion of the 2-year and 10-year yield spread on Treasury securities indicates that a recession will occur within 12 months. I’ve warned them that while the indicator has a good track record, it’s far from perfect.
This yield spread inverted on April 1st, 2022. In retrospect this inversion seems to have been an April fool’s joke, as yesterday’s jobs figures were extremely strong, showing payroll employment rising by 339,000 in May 2023. You don’t see those sorts of big job gains when the economy is in a recession. It looks like the “infallible” yield curve indicator has failed us.
In fairness, the 3-month/10-year spread didn’t invert until October of 2022, so that one might yet prove accurate. This inversion may have contributed to Bloomberg’s panel of forecasters telling us last October that a recession was 100% certain to occur by October 2023.
But there’s another problem with forecasts of an imminent recession. Bloomberg reports that a labor market showing this sort of very widespread strength almost never occurs within 6 months of the next recession:
Meanwhile, the so-called diffusion index showed that 60.2% of industries added jobs, a proportion that’s historically inconsistent with a recession starting in the next six months. That means that the majority of economists projecting economic contractions in the third and forth quarters of this year may have to reconsider. The labor market is often considered a “lagging indicator,” but it’s rare — outside of the highly unusual Covid-19 experience — for it to abruptly stop with no forewarning.
I don’t know what will happen between now and October, but I do know one thing for certain; one of these reliable recession indicators will be wrong. Either the yield curve inversion will predict a recession that didn’t happen, or the diffusion index will fail to give its normal advanced warning.
This is one more example of why you should always maintain a healthy skepticism when someone shows you a supposedly sure fire indicator of future economic trends. As they say in the investment industry, past performance is no guarantee of future success.
PS. It’s true that the yield spread accurately predicted a US recession in 2020. But in my view it just got lucky. Without Covid, I doubt whether a recession would have occurred during 2020.
