I’ve been irreverent at times about the yield curve as a recession predictor.

I instinctually recoil at the idea of infallibility, unless it applies to me, in which case I’m absolutely fine with it. (I’m just joking.)

Over the weekend, in “What If Bonds Don’t Rally In A Recession,” I tempered the sarcastic derision I sometimes employ when editorializing around the all-knowing curve. “The three-month, 10-year curve is at least as reliable as any other recession canary,” I wrote. If you accept the notion that it’s never failed to predict a downturn, it’s either in the process of being nine for nine or wrong for the first time.

Opinions (still) vary. Cam Harvey, not exactly an impartial observer, warned this month that markets ignore the signal from the curve “at their own risk.” BofA’s Michael Hartnett recently exhorted investors to “respect the lags,” a reference to the notion that the bottom could fall out for a previously resilient economy at any time given the rapidity and scope of the tightening cycle.

On Monday, JPMorgan analysts led by Marko Kolanovic likewise cautioned market participants against dismissing the curve. To wit:

The longer-term history of the yield curve inversion argues in favor of a bearish outcome, and we highlight three points. First, the long period required for core inflation to come down after peaks. Also for the assumed disinflation, we question the extent to which the disinflation will be immaculate, with concerns about company margins and profitability creating risk of a worsening job market. If the Fed funds rate is close to its peak, then history demonstrates that the probability of EPS downgrades is now exceptionally high, with layoffs needed to restore profitability. Second, it is difficult to have such a tightening of policy as we have had, via both rate hiking and balance sheet reduction, and not have a recession. If it has not materialized, then our view is that the lagged impact is coming. Third, it was easier to produce a soft landing if the rate hiking stopped before the yield curve inverts (early 60s, 80s and mid-90s), although we are well beyond that point in the current cycle. While the future is not written, we are not so quick to dismiss yield curve signals.

In the same note, Kolanovic and co. wrote that recent central bank communications, while differing in some respects, were consistent in echoing the “higher-for-longer” narrative.

That won’t likely end well, they suggested. “This signal aligns with our economists’ boiling-the-frog outcome, where resilience promotes sticky inflation and a need for sustained restrictive policy stances that, in turn, compress profit margins, erode balance sheet health and bring an end to the expansion,” Kolanovic’s team said.