U.S. macro monitor · September 22, 2026
U.S. Recession Risk: Growth Slows, Labor Holds
A slower second quarter is a warning to watch, not evidence that the expansion has ended. Labor remains the key confirmation test.
Executive Summary
August unemployment was 4.1%; the Sahm indicator was −0.07 percentage point, well below its 0.50-point trigger.
Real GDP rose in Q2, though at about 1.5% annualized from the published level series. Watch hiring and household demand for a joint turn.
A modest growth cushion makes the expansion more sensitive to renewed price shocks or tighter credit. This is a scenario judgment, not a modeled probability.
The clearest disagreement is between slowing output momentum and a labor market that has not crossed a recession threshold. The New York Fed’s smoothed recession-probability series stood at 0.76% in July; a Polymarket contract for a U.S. recession by the end of 2026 traded around 8.5¢ for Yes late on September 22. These measure different events and horizons, so neither number should be mistaken for our 3-, 6- or 12-month forecast.
Signals in Context
Labor: unemployment and Sahm reading
Monthly readings, March–August 2026. Sahm trigger: 0.50 percentage point.
Real GDP level: growth, but a slower step
Billions of chained dollars, seasonally adjusted annual rate; Q1 2025–Q2 2026.
Hover or tap charts for observations. GDP quarter-over-quarter annualized growth is calculated from adjacent published levels; economic releases can be revised.
Recession Risk Table
| Indicator | Latest reading | Interpretation | Portfolio implication |
|---|---|---|---|
| Sahm Rule (SAHMCURRENT) | −0.07 pp · Aug 2026 | Below trigger The 0.50 pp rule has not fired. | A labor-break thesis needs a sustained reversal, not one soft print. |
| Unemployment (UNRATE) | 4.1% · Aug 2026; 4.3% in March | Stable The recent move is down, not up. | Watch claims and payroll breadth before rotating fully into recession defenses. |
| Smoothed recession probability (RECPROUSM156N) | 0.76% · Jul 2026 | Low A lagged coincident recession signal, not a 12-month forecast. | Do not confuse a quiet current-state model with zero forward tail risk. |
| Real GDP (GDPC1) | $24.27T · Q2 2026; ~1.5% annualized q/q | Slower Output expanded, but less quickly than Q1 (~2.1% annualized). | Favor resilient balance sheets where earnings assumptions require less acceleration. |
| Money-market fund assets (MMMFFAQ027S) | $8.44T · Q2 2026 | Ambiguous Large cash balances are not a standalone recession signal. | Cash offers optionality; its size alone does not establish impending outflows from equities. |
| Recession search interest | 2,862 U.S. daily searches · Sep 21 | Sentiment About 19.7% above a week earlier on this series. | Attention can move faster than activity; use it as context, not a trading trigger. |
| 2026 recession event market | ~8.5% Yes · late Sep 22 | Market price Year-end contract, not a one-year economic forecast. | Thin event-market pricing and contract definitions warrant caution. |
Professional Commentary & Outlook
The apparent contradiction is less mysterious than it looks: GDP is a quarterly level, unemployment is a monthly household survey and the Sahm measure smooths labor-market changes. A smaller positive GDP step can coexist with a stable jobless rate. The Q2 GDP level rose from $24.18 trillion in Q1 to $24.27 trillion, a roughly 1.5% annualized gain; that is deceleration, not contraction. The Sahm reading, meanwhile, fell from 0.20 in March to −0.07 in August.
For the near term, the most informative failure mode is a linked deterioration: joblessness rising persistently, real consumption weakening, and credit spreads widening together. If only search interest or prediction-market pricing moves, it may reflect fear rather than a change in the economy’s operating rate. Conversely, a sustained hiring reversal would make the apparently comforting July probability reading stale.
For portfolio construction, distinguish a recession hedge from a bet on its timing. High-quality bonds can cushion a demand shock, but an inflation-driven selloff can hurt both stocks and long-duration bonds. Keep liquidity and position sizes aligned with drawdown tolerance; test cyclical earnings against slower revenue rather than assuming an abrupt recession. If labor and demand remain firm while output reaccelerates, paying heavily for defensiveness could create opportunity cost.
Sources and Definitions
Federal Reserve Economic Data: Sahm Rule, unemployment, smoothed recession probability, real GDP and money-market fund assets. Search figure: Daily Search Volume. Year-end contract and its resolution rules: Polymarket.
Money-market assets are converted from millions of dollars to trillions. Probability series and market price have different definitions; no single number here is a PortfolioAI recession-probability forecast. Information only, not personalized investment advice.