Inverted yield curve
Also called yield curve inversion, inverted curve, negative term spread
A yield curve on which short-term yields are above long-term ones, so the term spread is negative. In postwar US data, inversions have generally been followed by recessions, which is why the curve is watched as a leading indicator.

- Level
- Intermediate
- Reading
- 2 min
- Sources
- 5Central bankRegulator
- Updated
On this page· 5 sections
Lending money for ten years usually pays more than lending it for three months. When the longer loan pays less, the yield curve is inverted[3].
The shape has a long record of appearing before US recessions, which is why forecasters and policymakers watch it closely[4]. That record is history. It describes what followed past inversions, not what any single inversion means.
How it works
An inversion is a negative term spread: a long-term yield minus a short-term yield comes out below zero[1]. Which pair is used matters. Commentators typically quote 10-year minus 2-year Treasury yields, but in San Francisco Fed tests on monthly data from 1972 to 2018, the 10-year minus 3-month spread was the most useful for forecasting recessions[1].
The record, as the sources date it:
- From January 1955 to February 2018, every US recession was preceded by an inversion of the 10-year minus 1-year spread. A rule that predicted a recession within two years of an inversion caught all nine, with one false signal in the mid-1960s. The lag ranged from 6 to 24 months[4].
- The Cleveland Fed’s summary, as of November 2026, notes that the curve inverted in May 2019, almost a year before the recession that began in March 2020. It lists three false positives: an inversion in late 1966, a very flat curve in late 1998, and the inversion from late 2022 through late 2024[5].
In past episodes the spread fell mainly because short-term rates rose sharply as the Federal Reserve tightened, while long-term rates moved more slowly. A flatter curve also makes borrowing short and lending long less profitable for banks, which may tighten credit[4]. Cause and effect are unclear: tight policy may slow the economy, or expectations of a downturn may push investors into long-term bonds and lower their yields[1].
Example
Say the 3-month bill yields 5.00% and the 10-year note 4.20%.
- Spread: 4.20 − 5.00 = −0.80 point, or −80 basis points. The curve is inverted on this measure.
- In Estrella and Mishkin’s 1996 model, estimated on data from 1960 to 1995, a quarterly average spread of −0.82 point went with an estimated 50% probability of recession four quarters later[2]. This spread sits close to that level.
- In the 1955–2018 record, recessions began 6 to 24 months after the spread turned negative[4].
Step 3 describes past lags only. The inversion of late 2022 to late 2024 is one the Cleveland Fed lists as a false positive[5].
Watch out
An inversion is not a recession call. The record includes false signals, from the mid-1960s to 2022–2024[4][5].
Correlation is not causation. The San Francisco Fed warns that the predictive link says little about what causes recessions[1].
Model probabilities come from their sample. The 1996 estimates rest on data from 1960 to 1995[2], and later decades need not follow them.
Sources
- Central bankFederal Reserve Bank of San FranciscoInformation in the Yield Curve about Future Recessions (FRBSF Economic Letter 2018-20, M. Bauer and T. Mertens)Checked November 9, 2026
- Central bankFederal Reserve Bank of New YorkThe Yield Curve as a Predictor of U.S. Recessions (Current Issues in Economics and Finance 2(7), 1996, A. Estrella and F. S. Mishkin)Checked November 9, 2026
- RegulatorMunicipal Securities Rulemaking BoardMSRB Glossary of Municipal Securities Terms, Third Edition (August 2013)Checked November 9, 2026
- Central bankFederal Reserve Bank of San FranciscoEconomic Forecasts with the Yield Curve (FRBSF Economic Letter 2018-07, M. Bauer and T. Mertens)Checked November 9, 2026
- Central bankFederal Reserve Bank of ClevelandYield Curve and Predicted GDP GrowthChecked November 9, 2026
