Inverted Yield Curve

An inverted yield curve is short-term Treasury yields trading above long-term yields, and it has preceded every US recession since the late 1960s by 6 to 24 months. The 2022-2024 inversion was the longest on record; the signal is a calendar warning, not a stopwatch.

Macro & PolicyLast reviewed 2026-10-08

What it is

An inverted yield curve is the unusual state where short-term Treasury yields are higher than long-term yields, measured most often by the 2-year-versus-10-year spread going negative. Normal curves slope upward because lending longer carries more risk. An inversion means the market expects the short end to be cut, which historically means recession.

In the US the track record is the reason anyone watches it. Every recession since the late 1960s was preceded by a 2s10s inversion, with a lag of roughly 6 to 24 months. The mechanism is two-sided: the Fed sets the short end, and the market prices the long end on growth and inflation expectations. An inverted curve also squeezes banks, which borrow short and lend long. The signal is real; the stopwatch is not.

Key takeaways

An inverted yield curve is short-term yields above long-term, usually measured by the negative 2s10s spread.
Every US recession since the late 1960s was preceded by a 2s10s inversion, lagging 6 to 24 months.
The mechanism is two-sided: Fed-set short rates above market-priced long rates, plus a bank credit squeeze.
The 2022-2024 inversion ran roughly two years, the longest on record, and the recession it advertised arrived late or not at all.
The signal is a calendar warning for positioning, not a market-timing stopwatch.

How it works

The mechanism has two ends. The Fed sets the short end of the curve directly through the target range; the market prices the long end on expected growth and inflation over a decade. An inversion means the market is pricing rate cuts ahead: the long end falls below the short end because investors expect the short end to be coming down, and rate cuts historically arrive because recessions arrive first. The bond market is voting with its term structure.

The credit channel runs alongside. Banks borrow short (deposits, wholesale funding) and lend long (mortgages, business loans), so an inverted curve inverts their margins, and lending tightens; the tightening is the transmission from signal to recession. Money market funds also out-yield bond funds during inversions, which pulls cash up the curve and starves the long end further.

The timing is the part the models get wrong. The 2006 inversion preceded the December 2007 recession by roughly 22 months; the August 2019 inversion was followed by a pandemic recession in six months, which nobody curve-shaped; and the 2022-to-2024 inversion ran roughly two years, the longest on record, without a recession declared on schedule. The signal is a probability distribution, not a date.

Why it matters

The inverted curve is the bond-market recession signal with the best track record available, and it is also the only one that pays you to watch. During inversions, T-bills and money market funds out-yield bonds, which makes the short end the right place for cash by yield alone. The 2022 inversion had 12-month Treasuries paying above 5% while the aggregate bond index fell 13%. The short end was both the safe and the higher-yielding position.

I watched a colleague named Brent (yes, that Brent) move his entire portfolio to cash on inversion day. He sat there for two years through an 18% index year in 2023. The signal is a warning about the economy, not a sell order on the market, and the two are not the same instrument.

The highest-leverage thing most Americans can do is treat inversions as a positioning signal. Shorten duration for cash, keep the equity allocation written and untouched, and let the 6-to-24-month lag be the calendar rather than the stopwatch. Run the spread with the bond yield calculator and stress the allocation with the monte carlo calculator.

Real-world examples

The 2006-2007 inversion

The 2s10s spread inverted in 2006 (the Fed at 5.25%, the 10-year falling below it) and stayed inverted through 2007. The recession began in December 2007, roughly 22 months after the first inversion, and the S&P 500 peaked in October 2007, eleven months after the curve. The curve called the economy before it called the market, which is the normal order: the signal leads the recession by 6 to 24 months and the market by less. An investor who sold on inversion day sat through an 18% index gain in 2006-2007 before the fall.

The 2022-2024 record inversion

The 2s10s spread first inverted in April 2022, ran persistently from July 2022, and un-inverted in September 2024: roughly two years, the longest inversion on record. During the inversion, 12-month Treasuries paid above 5% while the aggregate bond index fell 13% in 2022, making the short end both the safe and the higher-yielding position. As of this writing, the recession the inversion advertised has not been declared on schedule, which makes it the most instructive false positive since the 1960s. The signal has a track record, not a guarantee.

The 2019 inversion, six-month lag

The 2s10s spread inverted in August 2019 for the first time since 2007, and the pandemic recession began in February 2020, six months later: the shortest lag on record, and one no curve-shaped forecast produced. The inversion had short Treasuries out-yielding the long end through late 2019, which made the money market fund the right cash position on yield alone. The lag is a distribution, not a date, and the 2019 tail proved it.

Common mistakes

Mistake: Selling the entire portfolio on inversion day.

Why it bites you: The signal leads recessions by 6 to 24 months and the market by less, and the 2006 inversion was followed by an 18% index gain before the fall. Brent inversion-day cash move sat out two years of 2023 recovery.

Mistake: Treating the inversion as a recession guarantee.

Why it bites you: The 2022-2024 inversion ran two years, the longest on record, without a recession declared on schedule. The signal is a probability distribution, and the false positives (1966, 1998 mid-cycle) are part of the record.

Mistake: Watching only the 2s10s spread.

Why it bites you: The 3-month-versus-10-year spread (the Fed preferred version) and the full curve matter, and the measures disagree at turning points. The 2s10s inverted months after the 3m10s in 2022, which changed the clock by half a year.

Mistake: Missing the yield opportunity the inversion hands you.

Why it bites you: During inversions, T-bills and money market funds out-yield bonds, making the short end the right cash position on yield alone. The 2022 inversion paid 5%+ on 12-month paper while bonds fell 13%.

Ivy's take

I spent a morning with the 2s10s history back to 1968 and the recession dates beside it so you do not have to. The yield curve is the only recession indicator with a track record long enough to be worth the screen space. The honest framing is a signal, not a stopwatch: the 2006 inversion led by 22 months, the 2019 one by six, and the 2022 one by none. That is a distribution I would not set a calendar by.

My rule is mechanical: inversion shortens the cash duration (T-bills over bond funds, which the yield spread justifies on its own). The equity allocation stays alone, because the curve is an economic signal and the market is a different instrument. Brent sold everything on inversion day and bought the 2023 recovery late. The curve called the economy; it never called his portfolio.

Related jargon, defined

2s10s spread: The 10-year Treasury yield minus the 2-year; the classic inversion measure.
3m10s spread: The 10-year minus the 3-month T-bill; the Fed preferred recession-signal measure.
Normal curve: Long yields above short yields, the upward slope that compensates for lending longer.
Term premium: The extra yield for holding longer maturities; what an inversion collapses.
Money market fund: The fund that out-yields bonds during inversions; the short-end cash position.
Sahm rule: The unemployment-based recession trigger; the labor-market cousin of the curve signal.

Frequently asked questions

What does an inverted yield curve mean?+

It means short-term Treasury yields are above long-term yields, which signals the market expects rate cuts ahead. Historically that expectation is recession-driven, and every US recession since the late 1960s followed an inversion.

How long after an inversion does a recession start?+

Roughly 6 to 24 months: the 2006 inversion led by about 22 months, the 2019 one by six. The lag is a distribution, not a date.

Has the yield curve ever given a false signal?+

Yes: the 1966 and 1998 mid-cycle inversions were not followed by recessions, and the 2022-2024 inversion (the longest on record) had no recession declared on schedule. The signal is a probability, not a guarantee.

Should I move my portfolio to cash when the curve inverts?+

The signal is about the economy, not the market, and the 2006 inversion was followed by an 18% index gain before the fall. Shortening cash duration makes sense (T-bills out-yield bonds); selling equities is a different decision.

What does an inverted curve mean for bonds?+

Short-term paper out-yields long-term bonds during inversions, which makes T-bills and money market funds the higher-yielding cash position. The 2022 inversion paid 5%+ on 12-month Treasuries while the bond index fell 13%.

Disclaimer: This information is general in nature and does not constitute financial or legal advice. Always consult a qualified professional for your specific situation.

Keep learning

Yield Curve Federal Funds Rate Recession Bonds Treasury Securities (T-Bills, Notes, and Bonds)

Inverted Yield Curve