PortfolioAI Macro Risk Monitor · August 21, 2026
Recession Outlook: Low Odds, Thin Margin for Error
The expansion remains intact on labor and output measures, but the margin for absorbing a renewed demand, credit or energy shock is narrower than the headline odds imply.
Executive Summary
Base case: continued expansion, not recession. July unemployment was 4.1% and the Sahm Rule reading was -0.03, comfortably below the 0.50 level commonly associated with a material labor-market deterioration. Real GDP rose to $24.27 trillion in the second quarter, while the smoothed recession-probability series was 0.6% in its latest reading. The core data therefore do not support a near-term contraction call.
The more relevant question is resilience. The weekly business-cycle index was last reported at 8.3 in May, below its February high, and money-market fund assets remained above $8 trillion in the latest quarterly data. Meanwhile, Polymarket pricing puts the chance of a recession by year-end near 8%. That is a tail-risk price, but it also argues for portfolios that can tolerate a slower landing rather than relying on a perfect one.
Editorial scenario assessment, not an econometric forecast.
Labor has not delivered a recession signal
Unemployment rate and Sahm Rule, through July 2026.
Output remains positive
Real GDP, seasonally adjusted annual rate, billions of chained dollars.
Low model risk, slower cycle impulse
Smoothed recession probability and the weekly business-cycle index use separate scales.
Recession Risk Scorecard
| Indicator | Latest reading | Signal | Interpretation |
|---|---|---|---|
| Sahm Rule (SAHMCURRENT) | -0.03 · Jul. 2026 | Low risk | Far below the 0.50 trigger; unemployment has not accelerated in the pattern associated with recession calls. |
| Unemployment rate (UNRATE) | 4.1% · Jul. 2026 | Stable | Below the 4.5% late-2025 high, preserving a meaningful household-income cushion. |
| Real GDP (GDPC1) | $24.27T · Q2 2026 | Expanding | Real output increased from $24.18T in Q1; revisions may alter the pace, not the current direction. |
| Smoothed recession probability (RECPROUSM156N) | 0.6% · Jun. 2026 | Low risk | The model remains in a calm range and far from readings usually associated with contraction. |
| Weekly business-cycle index (BCIG) | 8.3 · May 8, 2026 | Watch | Positive, but below February's 10.6. Its publication lag makes it both useful and incomplete as a current signal. |
| Money-market fund assets (MMMFFAQ027S) | $8.29T · Q1 2026 | Mixed | Large cash balances are a liquidity reserve and potential buying power, while also reflecting a durable preference for safety. |
| Polymarket: recession by end-2026 | ~8% · Aug. 21, 2026 | Tail risk | Prediction-market pricing remains below one chance in ten; it is a sentiment cross-check, not a standalone forecast. |
| U.S. searches for “recession” | 2,512 · Aug. 20, 2026 | Mixed | Search volume was 35.2% above roughly a week earlier and 39.6% above roughly a month earlier, but 53.5% below a year earlier. |
Where the Margin Gets Thin
Labor is the confirmation test
A renewed rise in unemployment and a Sahm move toward 0.50 would carry more weight than isolated payroll noise. Until then, the labor data argue against treating a recession as imminent.
Demand must keep the handoff intact
Consumer spending, business investment and credit availability need to reinforce one another. A synchronized weakening would matter more than a slowdown in any one interest-sensitive pocket of the economy.
Twelve months carries the wider range
The delayed effects of financing conditions, earnings deceleration and policy uncertainty widen the distribution of outcomes over a year, even when today’s recession probability is low.
Portfolio Implications
The evidence favors a resilient-growth posture with intentional downside diversification. Investors do not need to position for a recession as the base case; they do need to distinguish companies that can finance through a slower environment from those whose thesis depends on an uninterrupted acceleration in demand.
- Favor durability: recurring revenue, essential demand, pricing power and manageable refinancing needs remain the most useful quality filters.
- Keep defenses diversified: health care, utilities and high-quality duration can help offset a deterioration in labor or credit conditions without requiring a binary macro bet.
- Demand confirmation: a Sahm move toward 0.50, a sustained unemployment reversal and weaker cycle data together would justify a more defensive stance than headlines alone.
Sources and Notes
Horizon assessments are editorial scenario weights. Economic data are revised and published on different schedules.
Economic series: Federal Reserve Economic Data: SAHMCURRENT, BCIG, RECPROUSM156N, GDPC1, UNRATE, and MMMFFAQ027S. Search interest: Daily Search Volume. Event-market reference: Polymarket.