US Recession Indicators
Five widely-followed recession signals, each shown against its recognised threshold, from public federal data. Not a forecast: a scoreboard of what the best-known indicators are saying right now.
By Joey van Diest, founder and editorUpdated
Recession signals active
0 of 5
None of the tracked signals is above its recession threshold. That describes conditions today; it is not a promise about tomorrow. This is our editorial tally of well-known indicators, not a prediction or a combined index.
Sahm Rule
Signal: ≥ 0.50 pp
0.07 pp
not triggered
Since 2000 · shaded = NBER recessions · dashed = threshold · to Jun 2026
Yield curve (10y minus 3m)
Signal: inverted (< 0)
0.76%
not triggered
Since 2000 · shaded = NBER recessions · dashed = threshold · to Jul 2026
Smoothed recession probability
Signal: ≥ 20% (editorial)
0.5%
not triggered
Since 2000 · shaded = NBER recessions · dashed = threshold · to May 2026
Chicago Fed activity index
Signal: ≤ -0.70 (3-mo avg)
-0.05
not triggered
Since 2000 · shaded = NBER recessions · dashed = threshold · to Jun 2026
Initial jobless claims
Signal: > 10% above a year ago
187k
not triggered
-14.61% vs a year ago
Since 2000 · shaded = NBER recessions · dashed = threshold · to Jul 2026
Unemployment rate
Context (not counted)
4.2%
-0.1pp / yr
Since 2000 · shaded = NBER recessions · dashed = threshold · to Jun 2026
How to read this dashboard
No single indicator calls recessions reliably, which is exactly why traders watch a basket of them. Each one here is a genuine, published series with a threshold that has meaning in the record, and each fails in its own way: the yield curve is early and un-inverts before the downturn arrives; the Sahm Rule is prompt but confirms rather than predicts; the smoothed probability is a nowcast of where we already are. When several line up, the signal is stronger than any one alone. When they disagree, as they often do near turning points, that disagreement is itself the information.
Sahm Rule
Triggers when the 3-month average unemployment rate rises 0.50pp above its low of the prior 12 months. It has flagged every recession since 1970 with few false alarms, though it confirms a downturn rather than predicting one.
Yield curve (10y minus 3m)
The 10-year minus 3-month Treasury spread has inverted before every US recession since the 1960s, usually a year or more ahead. Its quirk: the curve often un-inverts shortly before the recession actually begins, so a return to positive is not an all-clear.
Smoothed recession probability
The Chauvet-Piger model infers the probability the economy is currently in recession from four coincident indicators (payrolls, income, production, sales). It sits near zero in expansions and spikes toward 100% in downturns; we mark 20% as an editorial "elevated" line.
Chicago Fed activity index
The Chicago Fed National Activity Index blends 85 monthly indicators into one gauge of US growth, where zero is trend. Its three-month average falling to -0.70 following an expansion has historically marked the onset of recession; positive readings mean above-trend growth.
Initial jobless claims
Weekly first-time unemployment filings are among the most timely labour signals. A sustained rise, we flag more than 10% above the year-ago level, has accompanied the onset of past recessions; claims are noisy week to week, so the trend matters more than any single print.
Unemployment rate
Shown for context rather than counted, since the Sahm Rule already distils it. The level matters less than the direction: unemployment turning up off a cycle low is the classic early tell, which is exactly what the Sahm Rule formalises.
The honest limitations
These indicators are coincident or lagging as often as leading, and thresholds calibrated on a handful of postwar recessions carry real uncertainty, seven or eight recessions is a small sample to draw bright lines from. The economy also changes: a signal that worked in an industrial economy may behave differently in a services-and-tech one. Treat the scoreboard as a disciplined way to check your assumptions against the data, not as a timing tool, and never as investment advice. It describes conditions; it does not tell you what to buy or sell.
Two of these signals have dedicated pages here. Theyield curve trackershows the 10y-2y and 10y-3m spreads live with more history, and thecentral bank rates dashboardshows the policy path that shapes the curve. New data lands on a fixed schedule; theeconomic calendarcarries the jobs and inflation releases that move these gauges.
Method and limitations
Every series is a US-government or public-domain series retrieved from FRED and baked into this page on our build schedule (these update monthly or weekly). Your visit reads a static copy, not an upstream API. The "signals active" tally is our editorial count of well-known indicators against their conventional thresholds; the 20% mark on the smoothed probability and the 10% mark on jobless claims are our own editorial lines and are labelled as such. NBER recession shading uses the official USREC series. This is descriptive, not a forecast and not investment advice.
This product uses the FRED® API but is not endorsed or certified by the Federal Reserve Bank of St. Louis.
Data sources
- Federal Reserve Bank of St. Louis (FRED) — the exact seven series charted: SAHMREALTIME, T10Y3M, RECPROUSM156N, CFNAIMA3, ICSA, UNRATE, USREC; public domain, monthly-baked
For general information and education only. This is not financial advice and not a recommendation to buy or sell anything. Trading and investing carry risk, including the risk of losing more than your initial outlay. Always verify figures against your broker or the original source before acting on them.
Spotted an error? Email[email protected]and it will be corrected. Maintained byJoey van Diest.