Yield Curve Explained: Shapes, Inversion, Recession Signal

TL;DR. The Treasury yield curve plots the interest rate on U.S. government debt against how long until that debt is paid back. Its shape — upward sloping, flat, or inverted — is one of the most-watched signals in finance, and an inverted curve has preceded every U.S. recession since the late 1960s. This piece explains what the curve is, how to read it, what its four common shapes mean, and where the signal can mislead.

What the Yield Curve Actually Is

The yield curve is a line connecting yields on otherwise identical bonds at different maturities. For the U.S. Treasury market, the Federal Reserve publishes daily constant-maturity yields in its H.15 release, from 1 month all the way out to 30 years. Because Treasuries carry no meaningful default risk for U.S.-dollar investors, the curve is treated as the baseline cost of money at each maturity, and almost every other dollar-denominated debt market — corporates, mortgages, munis, emerging-market sovereigns — prices off it.

Three things drive the level and shape of the curve:

Expected short-term rates. Long yields embed the path of expected future short rates set by the Federal Reserve.
Term premium. Lenders usually demand extra compensation for tying their money up for years — uncertainty about inflation, growth, and policy is greater the further out you go.
Supply and demand. Treasury issuance, foreign central-bank buying, bank balance-sheet needs, and pension demand all bend the curve.

The Curve Today: A Snapshot

Here is the curve as of June 22, 2026, straight from the Fed’s H.15 release:

Maturity	Yield (constant maturity)
1mo	3.66%
3mo	3.85%
6mo	3.98%
1y	4.04%
2y	4.24%
3y	4.25%
5y	4.29%
7y	4.39%
10y	4.51%
20y	4.97%
30y	4.95%

Source: Federal Reserve H.15 Selected Interest Rates, data for June 22, 2026.

Reading across, the curve is mildly upward sloping — about 129 basis points from the 3-month bill (3.85%) to the 30-year bond (4.95%). The “belly” of the curve (2- to 5-year) is unusually flat at 4.24% to 4.29%, and the 10-year/2-year spread sits at roughly 27 bp — positive, but historically thin. The picture below visualizes the same data.

The Four Common Shapes

Practitioners usually classify the curve into four canonical shapes. Each tells a different story about growth, inflation, and Fed policy expectations.

Normal (Upward Sloping)

The default shape: longer maturities yield more than shorter ones. This is consistent with positive expected growth, modest inflation, and a positive term premium. Most of post-war U.S. history looks like this.

Flat

Short and long yields are roughly equal. A flat curve often appears in transitions — late in a tightening cycle, when the Fed has pushed short rates up to meet long rates that are anchored by lower long-run inflation expectations. It is a “wait and see” shape.

Inverted (Downward Sloping)

Short yields are higher than long yields. Mechanically, the market is saying it expects the Fed to cut rates significantly in the future — which usually means it expects a recession. This is the shape that gets the most attention, and we will return to it in the next section.

Steep

Long yields are well above short yields. This typically shows up early in a recovery: the Fed has slashed short rates to support growth, and the long end is pricing in eventual normalization, higher inflation, or both. A “bull steepener” (short rates falling) is generally pro-growth; a “bear steepener” (long rates rising while shorts hold) often signals fiscal or inflation concerns.

Why Inversion Gets the Headlines

A long-running observation in U.S. data is that the 10-year Treasury yield falling below the 2-year (or below the 3-month bill) has preceded essentially every NBER-dated recession since the late 1960s. The New York Fed publishes a recession probability model based on the spread between the 10-year and 3-month yields; the Cleveland Fed publishes a related yield-curve-implied GDP growth indicator. Both are watched closely by market strategists.

The table below summarizes the last four cycles, using FRED’s 10Y minus 2Y series and NBER’s dated recessions:

10Y minus 2Y first inversion	NBER recession start	Approx. lead time
Dec 1988	Jul 1990	~18 mo
Feb 2000	Mar 2001	~13 mo
Feb 2006	Dec 2007	~22 mo
Aug 2019	Feb 2020	~6 mo
Apr 2022	None as of Jun 2026	—

Sources: FRED T10Y2Y (10Y minus 2Y Treasury spread); NBER US business cycle dates. Inversion dates are first sustained move below zero; lead times rounded.

Two things stand out. First, the lead time is variable — anywhere from about six months to roughly two years. Second, the 2022 inversion is the awkward case. The 10Y minus 2Y spread went negative in early-to-mid 2022 and stayed negative until late 2024 — the longest sustained inversion in the available data — yet no recession has been dated by NBER as of June 2026. That has prompted plenty of “this time is different” arguments around post-pandemic fiscal stimulus, balance-sheet quantitative tightening, and the Fed’s expanded toolkit.

Why Inversion Is a Signal in the First Place

Mechanically, the 10-year yield is roughly an average of expected future short rates plus a term premium. If the market expects the Fed to cut rates substantially over the next few years — which is what it would do in a recession — the path of expected future shorts falls below the current short rate, and the 10-year yield drops below the 2-year. The curve inverts because the market is pricing future cuts.

The information content is not magic; it is consensus expectations. Inversions also tighten credit conditions directly: banks fund short and lend long, so when the curve inverts, bank lending margins compress and credit growth tends to slow.

A Worked Example: Reading the Current Curve

Take June 22, 2026 (the table above). Two useful reads:

10Y minus 2Y = 4.51% − 4.24% = 0.27%. Positive but thin. The market is not pricing imminent deep cuts, but it is not pricing a long expansion either.
10Y minus 3M = 4.51% − 3.85% = 0.66%. This is the spread the NY Fed model uses, and it sits comfortably positive — a normal-ish reading.

The mild concavity at the long end — 20-year (4.97%) trading slightly above the 30-year (4.95%) — is also worth a note. This is a recurring quirk of the U.S. curve driven by relatively thin 20-year supply and liability-driven demand for the 30-year by pensions and insurers. It is not a signal in itself.

Where the Signal Misleads

Different spreads, different signals. The 10Y−2Y and the 10Y−3M can disagree. The 2022–24 episode saw both invert, but in some past cycles one inverted without the other.
Term premium changes alone can move the curve. If foreign central banks aggressively buy 10-year Treasuries, they can push the long yield down without any recession signal — exactly what some analysts argued happened in 2005–06.
Long, variable lags. The signal averages out over decades; in any single cycle, the lead time has ranged from six months to about two years. It is not a market-timing tool.
Quantitative easing distortions. Large central-bank Treasury holdings can compress the term premium, making the curve flatter than it would otherwise be.