If you’re inclined to suggest the recession crowd is treading dangerously close to Boy Who Cried Wolf territory, I won’t blame you.

The problem with recession forecasting is that because the next downturn is always just a matter of time, anyone predicting a recession isn’t actually saying much. The only way to make recession forecasts meaningful is to put expiration dates on them, something forecasters are understandably loath to do.

The same is true with market crashes (and also with the end of the world, by the way). They’re inevitable, but it’d be ludicrous to suggest the handful of unsavory characters who built the bear blog cottage industry from the ashes of the financial crisis were vindicated by Q1 2016’s deflation scare or by Q4 2018’s mini-bear market or by the fleeting global crash that accompanied the original pandemic lockdowns in March of 2020.

This is the problem I have with Jeremy Grantham. “Jeremy Grantham can spot market bubbles. Now he’s warning of an ‘ominous’ bust,” somebody called Julia Horowitz wrote for CNN in April. Since that article was published, the Nasdaq 100 is up 13%. Sure, Grantham predicted a Maurice Sendak-style “wild rumpus” just prior to last year’s bear market, but as I’m compelled to remind new investors every couple of years, if you had a nickel for every time Grantham suggested markets were likely to crash, you’d be at least several dollars richer.

Nobody wants to say this explicitly (preferring instead to frame the issue as a matter of “variable” leads), but the yield curve (Cam Harvey in rates form) as a recession predictor is a bit like Grantham as a bubble spotter. Sure, you can create charts that make them both look like Nostradamus, but the reality is less exciting. Curve inversions often reflect macro and policy conjunctures that typify (or are conducive to) downturns and Grantham’s bubble calls are predicated on conditions that anybody who’s ever seen a speculative mania would likewise identify as a possible bubble.

The problem is that because economic storm clouds don’t usually clear up as quickly as real ones, and because it’s notoriously difficult to identify “peak” mania, yield curve recession warnings are considered “in effect until recession” and we celebrate market crash calls as prescient even if the crash doesn’t occur until years (or even decades) later. I don’t know about anybody else, but to me that’s always seemed asinine. We don’t do that with weather. “A tornado watch is in effect for your area until there’s a tornado.” There has to be a time limit — an expiration date — on recession and market crash calls, otherwise they aren’t “calls.”

All of that said, recession canaries continue to pile up outside the US labor market, and I’d be remiss not to highlight them. The figure on the right below (it’s from BofA, but you can easily recreate it yourself) shows the three-month moving average of South Korean export growth. As you can see, it’s rarely this negative outside of recessions. This is another global demand harbinger, much like Taiwan industrial output and exports+, both of which are likewise depressed.

I should note that the YoY decline in South Korea’s exports was shallower than anticipated in May, but “better than expected” isn’t the same as good.

The figure on the left above shows BofA’s global earnings model which suggests further downside “consistent with global PMIs, financial conditions, Asian exports and the yield curve,” the bank’s Michael Hartnett said.

He also pointed to what I’ll call “tracking error” between ISM manufacturing and nonfarm payrolls, as illustrated below.

As you can see from the chart header, Hartnett thinks ISM is right. “[The] labor market is always a lagging indicator, but remains very strong in the post-COVID world driven by hoarding, low participation rates [and] maybe immigration,” he wrote. “Big fiscal stimulus is set to fade in H2 but [there’s] no recession until the ultra-low 3.4% unemployment rate starts to rise.”

He was writing prior to the release of May’s jobs report in the US, which showed the unemployment rate did indeed rise, and rather sharply at that. And yet, despite weakness in the household survey, the NFP headline, at 339,000, was robust, and now makes for an even starker juxtaposition with ISM manufacturing (which, by the way, is now lower than it was when Hartnett annotated his chart).

So, recession? Yes. Surely. But when? If it’s next year, and it starts from SPX 4,700, then I’m not sure it makes sense to say the bears were “right.”