TL;DR. The yield curve plots U.S. Treasury yields against their maturities. Its shape — upward-sloping, flat, or inverted — reflects the market’s collective bet on future short-term interest rates and the term premium investors demand for locking up money longer. An inversion has preceded every U.S. recession since 1955, but the lead time has ranged from six to 24 months, and the July 2022 inversion produced no recession at all before the curve un-inverted in September 2024.
What the yield curve actually is
The U.S. Treasury issues debt at maturities ranging from four weeks to 30 years. The yield curve is simply a line connecting the yields on those instruments, plotted by time to maturity on the x-axis. When people say “the yield curve,” they usually mean the curve built from Treasury constant-maturity yields, which the Federal Reserve publishes weekly in its H.15 release and the Treasury publishes daily on its own site.
Because U.S. Treasuries are considered essentially free of default risk, their yields are a clean read on two things:
- What investors expect short-term interest rates to average over the life of the bond.
- The extra compensation — the term premium — investors demand for tying up capital longer and bearing duration risk.
Everything else in the bond market prices off that curve. Corporate bonds, mortgages, and swap rates are quoted as a spread over the equivalent Treasury.
The three canonical shapes
Normal (upward-sloping). Long yields exceed short yields. This is what you would expect if investors demand more compensation for longer commitments and are neutral or positive about future growth. Most of the post-WWII era has looked like this.
Flat. Short and long yields are similar. Often a transition state as the market re-prices Fed policy — either the Fed is close to done hiking, or close to done cutting.
Inverted (downward-sloping). Short yields sit above long yields. It signals that the market expects the Fed to cut rates in the future, usually because it expects growth or inflation to weaken. Historically it has been the most reliable single leading indicator of U.S. recession.
The current curve (July 2026)
As of the week ending July 24, 2026, the U.S. Treasury curve is upward-sloping across most of the maturity spectrum. The 10-year yields 4.69% versus the 2-year at 4.33%, a spread of roughly +36 basis points. The very long end (20-year at 5.18% and 30-year at 5.16%) sits a hair above the 10-year — a small hump at the far end that is not unusual and reflects supply/demand dynamics for the longest-maturity Treasuries as much as any macro signal.
| Maturity | Yield (weekly average) |
|---|---|
| 1M | 3.80% |
| 3M | 3.96% |
| 6M | 4.08% |
| 1Y | 4.14% |
| 2Y | 4.33% |
| 3Y | 4.36% |
| 5Y | 4.43% |
| 7Y | 4.55% |
| 10Y | 4.69% |
| 20Y | 5.18% |
| 30Y | 5.16% |
Compared with mid-2020, when 10-year yields briefly touched 0.5%, today’s curve reflects the higher rate regime that followed the 2021–2023 inflation surge, the more recent jump in hike odds after the May 2026 PCE print of 4.1%, and market expectations of persistently positive real rates.
Why the shape moves
Two forces drive the curve:
1. The expected path of short rates. Under the expectations hypothesis, the long yield should approximately equal the average of expected future short rates over the life of the bond. If the market expects the Fed to hold at 4.5% for the next decade, the 10-year should trade near that level.
2. The term premium. Because holding a long bond exposes you to reinvestment risk, inflation risk, and mark-to-market volatility, investors typically demand extra yield for going out on the curve. When that premium is high (as in the early 1980s), the curve steepens even if short-rate expectations do not move. The New York Fed publishes a monthly ACM term-premium estimate.
An inversion means that expected future short rates plus the (usually small or negative) term premium are actually lower than today’s short rate. That is the market voting, with money, that the Fed will have to cut — often because a recession is coming.
A worked example: what a 2-year Treasury is really pricing
Suppose the 3-month T-bill yields 4.00% and the 2-year note yields 3.50%. If the term premium is roughly zero, the 2-year is telling you the market expects the average 3-month rate over the next two years to be 3.50%. Since today’s 3-month is 4%, the market is implying meaningful rate cuts over the horizon — otherwise no rational buyer would lock up money for two years at a lower yield than they could roll at short-term. Any time a longer maturity yields less than a shorter one, this arithmetic is what you are actually seeing.
What inversion has — and hasn’t — predicted
Research from the Federal Reserve Bank of San Francisco finds that “every recession over this period [1955–2018] was preceded by an inversion of the yield curve” and that “the delay between the term spread turning negative and the beginning of a recession has ranged between 6 and 24 months.”
But the record is not spotless. The 10y–2y spread inverted in July 2022 — the longest and deepest inversion in modern U.S. history — and stayed negative until early September 2024. No recession has been NBER-dated in that window. Some economists argue the signal was distorted by post-COVID demand-supply mismatches and quantitative tightening; skeptics like to quip that yield-curve inversions have “predicted nine of the past five recessions.”
| NBER peak | NBER trough | Duration (months) |
|---|---|---|
| January 1980 | July 1980 | 6 |
| July 1981 | November 1982 | 16 |
| July 1990 | March 1991 | 8 |
| March 2001 | November 2001 | 8 |
| December 2007 | June 2009 | 18 |
| February 2020 | April 2020 | 2 |
The takeaway: the curve is a strong leading indicator, but it is not a switch. It should be read alongside credit spreads, unemployment claims, real M2 growth, and the ISM manufacturing index — never in isolation.
Which spread should you watch?
- 10y minus 2y (T10Y2Y). The most-watched by traders and financial media.
- 10y minus 3-month (T10Y3M). The Fed’s preferred version. The classic 1996 Estrella & Mishkin New York Fed study used this one, and it tends to have a tighter historical link to recessions.
- Forward-looking spreads. The 3-month forward, 18 months ahead, minus the current 3-month is another Fed-favored version that focuses on near-term rate expectations.
There is no single “right” spread. In practice, inversion across multiple pairs is a stronger signal than one solitary inversion.
Common mistakes
- Treating inversion as a timing tool. With historical lead times of 6–24 months and a recent 26-month false signal, the curve tells you a lot about direction and much less about when.
- Ignoring the reason. An inversion driven by a growth scare (long rates falling) is different from one driven by a hawkish Fed (short rates rising). Both are cautionary, but the second often resolves without a recession if inflation cools.
- Reading only nominal yields. The real (inflation-adjusted) curve, visible through TIPS, sometimes tells a different story than the nominal curve.
- Confusing the very long end with the belly. A 20y/30y inversion (like the small one visible today) is often about long-bond supply and pension demand, not the economic outlook.
Related concepts and what to learn next
If you want to go deeper, the natural adjacent topics are:
- Duration and convexity — how bond prices respond to yield changes, and why a 30-year Treasury is much more rate-sensitive than a 2-year.
- The term premium — the New York Fed’s ACM model estimate is updated monthly.
- TIPS and breakeven inflation — the real yield curve versus the nominal one.
- Credit spreads — investment grade vs. Treasuries, high yield vs. Treasuries. When those blow out at the same time the curve inverts, the recession signal is much stronger.
- Yield to maturity, YTC, and YTW and the mechanics of individual Treasury securities.
Sources
- Federal Reserve H.15 Selected Interest Rates (current yield levels).
- U.S. Treasury Daily Yield Curve Rates.
- NBER U.S. Business Cycle Expansions and Contractions.
- Federal Reserve Bank of San Francisco, Economic Forecasts with the Yield Curve (2018).
- Estrella & Mishkin, The Yield Curve as a Predictor of U.S. Recessions, FRB New York, 1996.
- FRED, 10-Year Minus 2-Year Treasury Spread (T10Y2Y) and 10-Year Minus 3-Month (T10Y3M).
- Federal Reserve Bank of New York, ACM Term Premia.
Disclosure: This article was produced with AI assistance and reviewed before publication. It is for informational purposes only and is not investment advice.