Yield Curve Inversion: The Recession Signal Explained

Key Takeaways

• Yield curve inversion has preceded every U.S. recession in the postwar era; the 10Y-3M spread is the version used in the NY Fed's recession-probability model.
• The typical lag from inversion to recession is 6–24 months; equities can rise for a year or more after the curve first inverts.
• Re-steepening from deep inversion is not an all-clear, it is often the late-cycle warning as the Fed begins cutting at the short end.
• Campbell Harvey first documented the 10Y-3M recession signal in his 1986 Duke dissertation; the Estrella-Mishkin probit model formalized it in 1996.

What It Is

The Treasury yield curve shows the interest rate the US government pays to borrow at different maturities on a given day. The Treasury publishes daily par yield curve rates at 11 constant maturities (1-month through 30-year), estimated from market quotes collected by the New York Fed at 3:30 p.m.

The curve takes three canonical shapes:

• Normal (upward sloping). Longer maturities yield more than shorter ones. This is the usual state of the world because investors demand a term premium for locking money up longer.
• Flat. Short and long yields cluster at similar levels. Often a transitional phase.
• Inverted. Short yields exceed long yields. Rare. In the postwar era it has preceded every US recession.

Two inversion measures get the most attention: the 10Y-2Y spread (10-year minus 2-year Treasury yields) and the 10Y-3M spread (10-year minus 3-month). Both are published by the St. Louis Fed's FRED database.

The Intuition

Why would anyone accept a lower yield on a 10-year bond than on a 3-month bill? Only if they expect short rates to fall sharply in the future. Short rates fall when the Fed cuts, and the Fed cuts when growth and inflation disappoint. An inverted curve is therefore the bond market collectively voting that policy is too tight and a slowdown is coming.

There is a second channel. Banks fund themselves short (deposits, repo) and lend long (mortgages, business loans). Their profit lives in the spread. When the curve inverts, that spread collapses, new lending becomes unprofitable, credit tightens, and the slowdown arrives a year or so later. The inversion is not just a prediction; it is part of the transmission mechanism.

How It Works

The recession-signalling property comes from the long-run consistency of this pattern. Cleveland Fed research notes that inverted yield curves have preceded each of the last eight US recessions as dated by the NBER. Economist Campbell Harvey first documented the relationship in his 1986 Duke dissertation and 1988 academic paper, using the 10-year minus 3-month spread.

The New York Fed's official recession-probability model, popularised by Arturo Estrella and Frederic Mishkin in 1996, also uses 10Y-3M. It fits a probit regression of NBER recession indicators on the spread and publishes a monthly probability-of-recession-in-12-months series.

A practical decomposition of the 10-year yield helps show why inversion is informative:

10Y nominal yield = (expected path of short rates) + term premium

If the 10-year is below the 2-year, markets are pricing an average short rate over the next decade that is lower than current short rates. That can only happen if investors expect material cuts.

The signal is not instant. Inversion typically precedes recession by 6 to 24 months, with the median closer to 12 months. The curve usually re-steepens before the recession actually begins, often because the Fed has started cutting at the short end. That re-steepening is not an all-clear. It is the next stage of the same cycle.

Worked Example

Walk through the 2022-2024 cycle as a case study. The 10Y-2Y spread first inverted briefly in April 2022, re-steepened, then inverted persistently from July 2022 onward. The 10Y-3M spread, slower to react, inverted in October 2022. By mid-2023 the 10Y-2Y was around -100 basis points, the deepest inversion since the early 1980s.

A textbook reading: recession probable within 12-24 months. By early 2024 both spreads began to re-steepen as markets priced in Fed cuts. The 10Y-2Y turned positive in September 2024. A growth slowdown and rising unemployment materialised through 2024 and 2025, though whether that counts as a full NBER recession is still being debated.

The 1998 Long-Term Capital Management episode is the canonical false positive: the 10Y-3M briefly went negative in September 1998, but no recession followed within two years. The 2019 brief inversion is a partial false positive, recession did arrive in early 2020, but it was triggered by the COVID shock rather than the endogenous credit cycle the yield curve normally signals.

Common Mistakes

1. Reading inversion as an immediate sell signal. The lag from inversion to recession is 6 to 24 months. Equities have historically continued to rise for a year or more after the curve inverts. In 2006, for example, the curve inverted in early spring but the S&P 500 peaked only in October 2007. Selling the day of inversion is usually far too early.

2. Treating any brief inversion as meaningful. A single daily print below zero does not constitute inversion in the research sense. The NY Fed and Cleveland Fed models use monthly averages. A reliable signal requires the spread to stay negative for at least several weeks. Intraday or one-day dips are noise.

3. Ignoring the re-steepening. The curve often returns to positive slope shortly before a recession begins, because the Fed has started cutting at the short end faster than the long end can follow. Treating re-steepening as evidence that "the signal was wrong" is a common error. Historically, re-steepening from deep inversion has been the late-cycle warning, not an all-clear.

4. Conflating 10Y-2Y with 10Y-3M. The two spreads usually agree but can diverge near cycle turns. 10Y-3M is more sensitive to Fed policy because the 3-month bill tracks the policy rate tightly. 10Y-2Y embeds more market expectation about future policy. Academic research has found 10Y-3M slightly more accurate for 12-month recession prediction, while practitioners often quote 10Y-2Y because it is easier to watch in real time.

5. Ignoring term premium when interpreting a flat curve. A flat curve during a period of low or negative term premium (like the 2010s QE era) is less informative than a flat curve during a period of normal term premium. If the long end is artificially suppressed by central bank purchases, the signal weakens. NY Fed research on term premium estimates helps adjust for this.

Frequently Asked Questions

What does it mean when the yield curve inverts? An inverted yield curve means short-term Treasury yields are higher than long-term yields. This happens when bond markets collectively expect the Fed to cut rates, which it does when growth slows. The 10Y-2Y and 10Y-3M spreads going negative are the most watched inversion measures.

Has the yield curve ever inverted without a recession following? Yes, but rarely. The clearest false positive was September 1998 during the LTCM crisis: the 10Y-3M briefly went negative but no recession followed within two years. The 2019 inversion is a partial case, a recession did arrive in 2020, but it was triggered by COVID rather than the endogenous credit tightening the curve normally signals.

How long after inversion does recession typically hit? The lag from inversion to NBER-dated recession has historically been 6 to 24 months, with the median near 12 months. Equities have continued to rise for a year or more after the curve initially inverts, making inversion a directional signal, not a market-timing tool.

What does yield curve re-steepening mean? When the curve re-steepens from deep inversion, usually because the Fed has begun cutting at the short end, it does not mean the recession risk has passed. Historically, re-steepening from deeply negative territory has been a late-cycle warning, not an all-clear. The curve in 2022–2024 followed this pattern closely.

What is the difference between the 10Y-2Y and 10Y-3M spreads? The 10Y-3M spread is more sensitive to current Fed policy since the 3-month bill tracks the policy rate closely. Academic research, including Campbell Harvey's original work and the Estrella-Mishkin model, finds 10Y-3M slightly more accurate for 12-month recession prediction. Practitioners often quote 10Y-2Y because it is easier to track in real time and embeds more market expectations about future policy.

Sources

1. Federal Reserve Bank of New York. "The Yield Curve as a Leading Indicator." https://www.newyorkfed.org/research/capital_markets/ycfaq
2. Federal Reserve Bank of Cleveland. "Yield Curve and Predicted GDP Growth." https://www.clevelandfed.org/indicators-and-data/yield-curve-and-predicted-gdp-growth
3. Federal Reserve Bank of San Francisco. "Current Recession Risk According to the Yield Curve." https://www.frbsf.org/research-and-insights/publications/economic-letter/2022/05/current-recession-risk-according-to-yield-curve/
4. Harvey, Campbell R. "Yield Curve Inversions and Future Economic Growth." https://people.duke.edu/~charvey/Term_structure/Harvey.pdf
5. U.S. Department of the Treasury. "Daily Treasury Par Yield Curve Rates." https://home.treasury.gov/resource-center/data-chart-center/interest-rates/TextView?type=daily_treasury_yield_curve

Disclaimer

This article is educational content only and is not financial advice. Nothing here is a recommendation to buy, sell, or hold any security. Consult a licensed advisor before making investment decisions.