Does an Inverted Yield Curve Mean a Recession Is Coming?
Direct answer
Yield curve inversions and recessions
Not necessarily. An inverted yield curve, when 2-year Treasury yields rise above 10-year yields, has come before every US recession since 1980, usually by 14 to 34 months. But the longest inversion on record, from mid-2022 to September 2024, has not been followed by an NBER-dated recession as of November 2026. It is a warning sign with a long and variable lag, not a timer.
- 10-year minus 2-year spread was +0.32 points on Nov 9, 2026 (FRED T10Y2Y)
- Inversions began 14–34 months before the 1980, 1990, 2001 and 2008 recessions
- The 2020 recession followed only a brief inversion in 2019
- The 2022–2024 inversion (low −1.08 points) has had no recession so far
The yield curve is one of the few recession signals that economists, bond traders and people on X all watch. Its record is impressive, which is exactly why the last few years have confused so many people.
What people on X are asking
In February 2026 a thread began with the line the yield curve just inverted, cue the recession headlines, but what does it actually mean? In April another explainer asked what the yield curve is and why it matters, and in September a finance series asked what the yield curve actually shows. Note that the daily FRED data below show the 10-year minus 2-year spread positive throughout 2026, so the February post may have referred to a different pair of maturities.
The fact-checked answer
What inversion means. Normally investors demand a higher yield to lend for 10 years than for 2. When the 2-year yield rises above the 10-year, the curve is "inverted". It usually happens when the Federal Reserve has pushed short-term rates up and bond markets expect slower growth and lower rates later.
UXS chart
US 10-year minus 2-year Treasury yield spread since 1976
10-year minus 2-year spread (percentage points): latest
0.32%
November 9, 2026
Highest in chart period
2.91%
February 4, 2011
Lowest in chart period
-2.41%
March 20, 1980
Change over 1 year
-0.24 pts
vs November 7, 2025
Year by year: 10-year minus 2-year spread (percentage points)
Computed by UXS from the same data as the chart. Tap a column header to sort. * = year to date (latest data November 9, 2026).
| 2026* | 0.32% | -0.39 pts | 0.74% | 0.20% |
| 2025 | 0.71% | +0.38 pts | 0.73% | 0.20% |
| 2024 | 0.33% | +0.68 pts | 0.33% | -0.47% |
| 2023 | -0.35% | +0.18 pts | -0.13% | -1.08% |
| 2022 | -0.53% | -1.32 pts | 0.89% | -0.84% |
| 2021 | 0.79% | -0.01 pts | 1.59% | 0.72% |
| 2020 | 0.80% | +0.46 pts | 0.83% | 0.12% |
| 2019 | 0.34% | +0.13 pts | 0.34% | -0.04% |
| 2018 | 0.21% | -0.30 pts | 0.78% | 0.11% |
| 2017 | 0.51% | -0.74 pts | 1.30% | 0.51% |
| 2016 | 1.25% | +0.04 pts | 1.34% | 0.76% |
| 2015 | 1.21% | -0.29 pts | 1.77% | 1.19% |
| 2014 | 1.50% | -1.16 pts | 2.61% | 1.46% |
| 2013 | 2.66% | +1.13 pts | 2.66% | 1.46% |
| 2012 | 1.53% | -0.11 pts | 2.00% | 1.21% |
The track record. We found every stretch when the spread was below zero for at least about three weeks (merging episodes less than six months apart) and matched it with the start of the next recession in the NBER-based FRED indicator.
Yield-curve inversions and the next recession (10y–2y spread, FRED daily data)
Computed by UXS from FRED T10Y2Y and USREC. The 1978–1982 episode spans the 1980 and 1981–82 recessions. A brief inversion in August 2019 (a few days) came before the 2020 recession, which was triggered by the pandemic. Tap a column header to sort.
| Aug 1978 | Jul 1982 | −2.41 | Feb 1980 | 17 |
| Dec 1988 | Mar 1990 | −0.45 | Aug 1990 | 19 |
| May 1998 | Jul 1998 | −0.07 | Apr 2001 | 34 |
| Feb 2000 | Dec 2000 | −0.52 | Apr 2001 | 14 |
| Dec 2005 | Jun 2007 | −0.19 | Jan 2008 | 24 |
| Apr 2022 | Sep 2024 | −1.08 | None as of Nov 2026 | – |
Why the signal can work. The New York Fed's research has found that the spread between long and short rates has predicted recessions two to six quarters ahead. Higher short rates squeeze bank lending margins and borrowing, and an inverted curve also reflects investors expecting the Fed to cut later.
Why it may have failed this time (so far). The 2022 inversion came from the fastest Fed tightening in decades, while the economy still had strong jobs growth and large household savings. Many economists also point to lower term premiums after years of bond buying by central banks. None of these explanations is proven, and a late recession is still possible.
The re-steepening. The spread turned positive in September 2024 and was +0.32 points on November 9, 2026. Historically, recessions often began after the curve had already un-inverted, because the Fed was cutting short rates by then. Our fed funds vs 10-year chart shows that pattern from the rates side.
The level of yields matters as well as the curve; see how rising 10-year yields affect the stock market.
Credit markets give a second warning signal; see what widening high-yield spreads say about stocks.
For what happened to stocks once recessions arrived, see how much the Nasdaq fell in each US recession since 1971.
Decision helper
| Signal | What it suggests | Limits |
|---|---|---|
| Curve inverted | Tight policy; markets expect cuts | Lead time of 1–3 years; can be a false alarm |
| Curve re-steepening after inversion | Fed easing; recession risk historically elevated | Also happens in soft landings |
| Curve steep and rising | Growth or inflation expectations rising | Can hurt bond prices |
| Other indicators weakening (jobs, claims) | Confirms the curve | Data revisions |
Checklist for reading the yield curve
- Say which spread you mean: 10y–2y and 10y–3-month can disagree.
- Look at how long and how deep the inversion is, not one day.
- Check jobless claims, payrolls and credit spreads alongside it.
- Remember the long lag; markets have often risen after inversions began.
- Avoid all-or-nothing portfolio moves based on one indicator.
Quick poll
What do you think the yield curve is telling us in 2026?
What each answer means
- A recession is still coming: The historical lags were up to almost three years, and past recessions often started after re-steepening. The data cannot rule this out.
- The signal broke this time: The 2022–2024 inversion has not been followed by a recession so far. It may simply be late, or the link may have weakened.
- It never mattered for stocks: Stocks have often risen for a year or more after an inversion began, but recessions that followed still hit stocks hard.
- I do not follow it: That is fine for long-term index investors; it is one indicator among many.
Your answer is stored only in this browser (localStorage). UXS does not collect, count or publish votes, so no totals are shown.
Common mistakes
- Treating inversion as a sell signal for stocks on the day it happens.
- Mixing up maturities and comparing different spreads.
- Ignoring that recessions are dated by the NBER months after they start.
- Assuming "this time is different" or "it always works" with certainty.
FAQ
What is an inverted yield curve?
When short-term government bond yields are higher than long-term ones, most often measured as the 10-year minus 2-year Treasury yield going below zero.
How long after inversion does a recession start?
In the five cases since 1978 in our table, 14 to 34 months after the inversion began.
Is the yield curve inverted now?
No. The 10-year minus 2-year spread was +0.32 percentage points on November 9, 2026.
Did the 2022 inversion predict a recession?
Not so far. It lasted from about April 2022 to September 2024, and the NBER had not dated a recession as of November 2026.
Why does the Fed matter for the curve?
The Fed sets short-term rates, which drive the 2-year yield. Rapid hikes push the short end above the long end.
Which spread is best?
Research by the New York Fed often uses the 10-year minus 3-month spread. FRED publishes both (T10Y2Y and T10Y3M).
Sources
- FRED: 10-year minus 2-year Treasury spread (T10Y2Y)
- FRED: NBER-based recession indicator (USREC)
- New York Fed: The yield curve as a leading indicator
- NBER: US business cycle expansions and contractions
Facts checked November 10, 2026. X posts are cited as evidence of the question, not of the answer.