Five US recession indicators compared
No single series calls a US recession — the NBER does, often a year late. These are the five indicators used to guess in the meantime, and each has failed at least once in a way worth knowing about.
At a glance
| Measure | Latest | Period | Published by | Frequency |
|---|---|---|---|---|
| Sahm Rule (real time) | −0.07% | Aug 2026 | Federal Reserve Bank of St. Louis / Claudia Sahm | Monthly |
| Recession probability (Chauvet-Piger) | 0.76% | Jul 2026 | Chauvet and Piger | Monthly |
| Treasury yield spread (10y minus 3m) | 0.93% | Sep 25, 2026 | US Treasury | Daily |
| Conference Board Leading Economic Index | 99.5 | Aug 2026 | The Conference Board | Monthly |
| CFNAI three-month moving average | 0.01 Standard deviations | Aug 2026 | Federal Reserve Bank of Chicago | Monthly |
How each one is built
- Sahm Rule (real time)
- Method: Three-month average unemployment rate minus its lowest three-month average of the past year
- Covers: Triggers at 0.50 percentage points
- Released: With the jobs report
- Recession probability (Chauvet-Piger)
- Method: Markov-switching model on payrolls, industrial production, income and sales
- Covers: Probability the economy is in recession now, 0–100%
- Released: With a lag of about two months
- Treasury yield spread (10y minus 3m)
- Method: Long yield minus short yield; negative = inverted curve
- Covers: Every US recession since 1969 was preceded by an inversion
- Released: Next business day
- Conference Board Leading Economic Index
- Method: Ten components including new orders, claims, building permits and the yield spread
- Covers: Composite, 2016 = 100
- Released: Third week of the following month
- CFNAI three-month moving average
- Method: 85 indicators reduced to one; zero = trend growth
- Covers: A reading below −0.70 has historically signalled recession
- Released: About four weeks after the month
What each one actually claims
The Sahm Rule and the Chauvet-Piger probability are COINCIDENT: they tell you a recession has probably already begun. They are not forecasts, and Claudia Sahm has repeatedly said her rule was designed to trigger fiscal support quickly, not to predict anything.
The yield spread and the LEI are LEADING, with long and variable lags. The curve has inverted before every US recession since 1969, but the gap between inversion and recession has ranged from about six to twenty-four months, which makes it close to useless for timing.
The CFNAI sits in between. It is a summary of current activity across 85 series, so it describes the present accurately and turns at roughly the same time as the economy does.
Which one to use
Use the Sahm Rule when you want a single, transparent, hard-to-game trigger on the labour market. Its weakness is that a rise in unemployment driven by more people entering the labour force looks the same as one driven by job losses.
Use the yield spread to judge whether policy is restrictive, not to date a downturn. Use the LEI for breadth: its component diffusion matters more than the headline, because a decline driven by one volatile input is far weaker evidence than a broad-based one.
For the official answer, there is only the NBER Business Cycle Dating Committee, which looks at depth, diffusion and duration together and announces months or years after the fact.
Frequently asked questions
- What triggers the Sahm Rule?
- It triggers when the three-month average unemployment rate rises 0.50 percentage points or more above its lowest three-month average of the previous twelve months.
- How far ahead does an inverted yield curve signal a recession?
- Historically between about six and twenty-four months, which makes it a reliable signal of direction but a poor guide to timing.
- Who officially declares a US recession?
- The Business Cycle Dating Committee of the National Bureau of Economic Research, usually many months after the turning point it identifies.
- Is two quarters of negative GDP growth a recession?
- That is a common rule of thumb but not the US definition. The NBER considers depth, diffusion and duration across employment, income, production and sales.
