Five dials with their trigger rules in plain English, a six-signal composite backtested to 1970, and the financial-conditions backdrop: whether money is loose or tight is the transmission belt from Fed policy to the real economy, and tight conditions slow growth and inflation with a lag.
Recession risk is read here from 6 signals, each with a stated trigger rule, and from a composite that counts how many of them are firing. The composite for September 2026 is 0 — the share of inputs triggered (4 of 6 inputs had data for the month); the series runs monthly since January 1971 (669 months), and has been at 50 or above in 63 of them, most recently June 2020.
| Signal | Latest | As of | Trigger rule | Status | Record since 1970 |
|---|---|---|---|---|---|
| Sahm Rule (real-time) | 0.00pp | September 2026 | ≥ +0.50pp | clear | 0 of 2 fresh triggers were followed by a recession within 12 months (0%) |
| 10Y–3M Yield Curve | 0.9 | September 2026 | < 0 (inverted) | clear | 4 of 9 fresh triggers were followed by a recession within 12 months (44%) |
| Chicago Fed NFCI | -0.53 | September 2026 | > 0 (tight) | clear | 4 of 10 fresh triggers were followed by a recession within 12 months (40%) |
| Claims Trend (3m/12m) | 0.954 | September 2026 | 3m avg 10% above 12m | clear | 1 of 2 fresh triggers were followed by a recession within 12 months (50%) |
| CFNAI (3-month MA) | 0.01 | August 2026 | < −0.70 | clear | coincident — fires inside recessions, not before |
| Recession Probability Model | 0.62% | August 2026 | > 20% | clear | 2 of 2 fresh triggers were followed by a recession within 12 months (100%) |
backtest: share of fresh triggers followed by an NBER recession within 12 months, 1970→. The "record" column only counts fresh triggers that fired outside a recession: the Sahm rule and the CFNAI are coincident gauges that confirm a downturn has begun rather than predicting it, so their lead record is not the point; the yield curve and the claims trend are the leading ones, at the cost of false alarms.
Sahm rule: the three-month average unemployment rate minus its low of the prior twelve months; ≥ +0.50 points has marked the start of every recession since 1970. Yield curve (10-year minus 3-month): below zero (inverted) has preceded every modern recession by 6–18 months; the spread is +0.83pp (October 16, 2026, FRED T10Y3M). Claims trend: the three-month average of initial claims 10% above the twelve-month average. CFNAI three-month average below −0.70. Recession probability (the Chauvet–Piger smoothed model on coincident data) above 20%; it reads 0.62% for August 2026. The Weekly Economic Index, scaled to GDP growth, is 3.05 (week of October 10, 2026); below zero is a contraction pace.
Conditions measure whether money is loose or tight — the credit-and-liquidity backdrop through which Fed policy reaches the real economy. On the Chicago Fed's NFCI zero is the historical average, positive is tighter than average (harder to borrow, wider spreads, less liquidity) and negative looser; it reads -0.5 (October 9, 2026). The adjusted index (ANFCI) removes the part of tightness explained by growth and inflation themselves: -0.49. The St. Louis Fed's financial stress index is centered on zero as normal; -0.72 today. Tight conditions slow growth and inflation with a lag, which is why they sit under the recession dials: a Gauge at 3.45% against an official CPI of 3.7% with loose conditions is a different problem from the same prices with tight ones. The consumer end of the same chain is the stress index; the policy rate driving it is on the rates and Fed Watch pages.