US 2Y/10Y Yield Curve Spread (T10Y2Y)
The US 10-year minus 2-year Treasury yield spread (T10Y2Y) from June 1976 to the latest reading (+0.51 points on 7 October 2026). The most-watched recession indicator in markets: every US recession since the series began was preceded by an inversion, with lags of roughly 6 to 24 months. The 2022-2024 inversion was the longest continuous one in the series, at just over 25 months. Includes a zero line so inversions stand out. Refreshed from FRED twice a day, with AUD-trader framing on how US recession risk reaches the ASX 200 and AUD.
Chart
10Y minus 2Y spread in percentage points. Below zero = inverted = recession warning. The 2022-2024 episode was the longest continuous inversion in the series, which starts in June 1976. Recessions shaded from 2001.
What is the 2Y/10Y spread?
Mechanically: 10Y yield minus 2Y yield, in percentage points. A positive value means long bonds yield more than short bonds (normal upward-sloping curve). A negative value means the inverse (downward-sloping or inverted curve).
The bond market expects, in equilibrium, to be paid more for taking 10 years of duration risk than 2 years. When this expectation reverses, it means the market believes short rates will fall significantly within the next 10 years - which historically happens because the Fed cuts rates in response to recession.
Historical 2Y/10Y inversions and US recessions
| Inversion start | Recession start | Lag | Deepest inversion |
|---|---|---|---|
| Aug 1978 | Jan 1980 | 17 mo | -2.41% |
| Sep 1980 | Jul 1981 | 10 mo | -1.70% |
| Dec 1988 | Jul 1990 | 19 mo | -0.45% |
| May 1998 (brief) | No recession | - | -0.07% |
| Feb 2000 | Mar 2001 | 13 mo | -0.52% |
| Dec 2005 | Dec 2007 | 24 mo | -0.19% |
| Aug 2019 (brief) | Feb 2020 (COVID) | 6 mo | -0.04% |
| Jul 2022 (brief dip Apr 2022) | None declared as of Sep 2026 | - | -1.08% |
The 2022-2024 inversion is the only sustained one in the table that has not been followed by an NBER-dated US recession. Some analysts argue the recession risk is still active given typical lags. Others argue the soft-landing thesis is intact and the historical pattern broke for structural reasons (fiscal dominance, COVID-era savings, AI capex).
Trader takeaway
- Inversion alone is not a sell signal. Equity returns in the 12 months after the initial inversion have often been positive. The recession risk tends to crystallise later.
- Re-steepening from inversion is more dangerous. When the curve normalises from inverted back toward zero or above, the Fed has typically already started cutting. In 1990, 2001, 2007 and 2020 the recession began 3 to 8 months after normalisation. The 2024 normalisation has broken that pattern so far.
- BTC AUD reaction. BTC has shown mixed behaviour through inversions: 2007-2008 inversion preceded the early BTC era; 2019 inversion preceded the March 2020 COVID crash and the rally that followed (BTC AUD rose more than twelvefold from the March 2020 low to the November 2021 high); 2022 inversion coincided with the BTC bear and 2024-2026 recovery.
- AUD/USD reaction. AUD/USD tends to weaken when US recession fears intensify (USD safe-haven bid) and recover as cuts arrive. Position size accordingly.
Methodology
- Source. FRED series T10Y2Y (computed by FRED as DGS10 minus DGS2 daily).
- Endpoint. Public fredgraph.csv.
- Recession shading. NBER-dated US recessions from 2001 onwards (2001, 2007-09, 2020). Earlier recessions are not shaded.
- Zero line. Threshold marker for inversion vs normal.
- Refresh cadence. An automated cloud refresh pulls the series from FRED twice a day (07:30 and 13:30 Sydney time) and the site rebuilds when the data changes. If FRED is unreachable, the last good data is kept.
Related tools
- 10Y Treasury Yield - the long-end input.
- 2Y Treasury Yield - the short-end input.
- Fed Funds Rate - the policy lever.
- US Unemployment Rate - the labour-market recession indicator.
- Broad USD Index - the safe-haven currency channel that transmits US recession risk to AUD/USD.
Frequently asked questions
It is the difference between the 10-year US Treasury yield and the 2-year US Treasury yield. Normally positive (long bonds yield more than short bonds because of term premium and growth expectations). Occasionally negative ('inverted'), meaning short rates exceed long rates - which has historically signalled that the Fed has tightened policy enough to slow the economy meaningfully. FRED publishes the spread directly as T10Y2Y.
The bond market is forward-looking. When the Fed has hiked aggressively, the 2Y reflects current restrictive policy at face value; the 10Y reflects the average expected short rate over 10 years. If the 10 is below the 2, the market is pricing the Fed cutting rates significantly within the next decade - which happens in response to a recession. All six US recessions since the T10Y2Y series began in 1976 were preceded by a 2Y/10Y inversion. The lag from the start of inversion to the start of recession ranged from about 6 to 24 months, averaging roughly 15. Studies using other maturity pairs trace the same pattern back to the 1950s.
Not as of September 2026: no NBER-dated recession has followed. The 2022-2024 inversion was the longest continuous inversion in the T10Y2Y series (just over 25 months, July 2022 to August 2024) and the deepest since September 1981 (minus 1.08 points on 3 July 2023). Two common explanations: (a) fiscal stimulus (CHIPS Act, Inflation Reduction Act, Infrastructure Act) offset monetary tightening; (b) the Fed managed a soft landing. The curve turned positive in late August and early September 2024, but that alone does not clear the risk: in 1990, 2001, 2007 and 2020 the recession began 3 to 8 months after the curve had turned positive again.
Through global recession-risk transmission. A US recession slows Australian growth through trade, commodity prices and financial conditions. AUD/USD tends to weaken on USD safe-haven flows. The ASX 200 has usually fallen with the S&P 500 in US risk-off episodes, typically by less. BTC AUD sold off hard in the March 2020 recession shock and through the 2022 tightening cycle before recovering. Watching the US curve is one of the highest-information macro signals for AU-resident portfolio construction.
Yes. The 3-month / 10-year (T10Y3M) is preferred by some Fed researchers because the 3-month is a closer proxy to Fed policy than the 2Y. The 3M/10Y curve also inverted in 2022 and signalled the same recession risk. The Fed's NY-research probit model uses 3M/10Y. Markets watch the 2Y/10Y more because it is easier to trade and more liquid. Both curves carry the same fundamental signal.
FRED series T10Y2Y, which is computed by FRED as DGS10 minus DGS2 daily. It starts in June 1976 and the chart shows the full history. An automated refresh pulls it from FRED twice a day (07:30 and 13:30 Sydney time) and the site rebuilds when the data changes; if FRED is unreachable, the last good data is kept.
About the author
