The 2022 Inversion Promised a Recession. 2024 Didn't Deliver One.
Yield-curve inversion has long been treated as the closest thing to a recession alarm. After the large and persistent inversion of 2022, most term-spread measures — 10-year minus 3-month, 10-year minus 2-year — flagged a high likelihood of recession in 2023-24. No recession arrived in 2024, and that miss has forced a reappraisal of the measure's reputation as the most reliable US recession predictor. A new paper by Laurent Ferrara and coauthors, “Missing Recessions: Revisiting the Term Spread's Predictive Power,” reports that adding the debt service ratio to the model reconciles the false alarm for both the US and Germany, following the approach of Borio and colleagues (2020). That leaves an open interpretive question: does the curve carry information of its own, or does it largely stand proxy for the debt burden that does the real work?
AI-generated discussion · Different perspectives
So the curve may never have been the alarm — the debt service ratio was the thing actually ringing.
Maybe not either/or. The curve could signal that growth is slowing, while the debt-service burden decides whether slowing turns into an actual recession.
Either way, anyone who wrote off the term spread entirely after 2024 is over-correcting from a single miss.