From July 2022 the 2-year/10-year Treasury curve was continuously inverted — eventually passing 624 days to become the longest inversion on record, exceeding the 1978 run. The 10Y–3M probit model's 12-month recession odds peaked near 71% — its highest reading since the early 1980s. Every recession playbook said the downturn was due in 2023. It never came.
Why the signal failed
The inversion's mechanism — inverted curves squeeze bank lending margins, choking credit — was blunted by a deposit glut: banks were paying near-zero on deposits, so margins stayed healthy and credit kept flowing. Meanwhile fiscal stimulus, locked-in low mortgages, and a labor supply shock kept demand resilient. The curve measured rate expectations accurately; it was the transmission channel that broke.
What it teaches
A single-signal monitor would have flashed red for two years straight and been wrong the whole time — the strongest possible argument for a composite panel. Credit spreads stayed calm, NFCI never sustained stress, and labor demand cooled without collapsing. The episode is included here deliberately: a monitor that only publishes its hits is marketing, not methodology.