Recession Risk 34/100 — September 2, 2026

Recession Risk Score: 34/100 — MODERATE (+0 vs 30 days ago)

Today’s Recession Risk Score holds at 34/100 (MODERATE), unchanged versus 30 days ago (Aug 3 → Sep 2). The macro picture still argues against an “imminent” recession because the highest-weight labor trigger set (Sahm Rule + claims) remains firmly unbroken. At the same time, the score refuses to fall into “low risk” because forward-looking cyclicals (temporary help, freight) and confidence keep flashing yellow-to-red. Net: the economy looks late-cycle but still expanding, with policy and confidence shock risk now the dominant tail.

Score Trend — Last 30 Days

Over the last 30 days (2026-08-03 → 2026-09-02), the score started at 34, ended at 34 (Δ 0), with a min of 34, max of 42, and average of 36. That profile matters: the system is not trending worse, but it did experience episodic stress spikes that failed to persist.

The “shape” looks mean-reverting rather than accelerating. We saw brief jumps to the low-40s (risk-on markets wobble / policy messaging tightening / soft data rolling over), followed quickly by snap-backs to 34 as hard labor and credit stayed resilient. In practice, this is what a late-cycle expansion often looks like: soft patches and volatility, but no sustained deterioration in the few indicators that typically break first before recessions.

Key Drivers

1. Sahm Rule remains untriggered (July 2026: -0.03)
   The Sahm Rule is still comfortably below the 0.50 trigger—a major anchor keeping near-term recession odds capped. With the July reading at -0.03, the unemployment upshift required for a rapid recession repricing simply isn’t present in the data yet.

2. Initial jobless claims remain exceptionally low (203k, week ending Aug 22; reported Aug 27)
   Claims at 203,000 reinforce the “layoffs still scarce” signal—one of the most reliable near-term recession off-ramps. This is consistent with a labor market that may be cooling at the margin but is not breaking. (apnews.com)

3. Yield curve re-steepening reduces “imminent recession” signaling (2s10s ≈ +40 bps)
   The 2s10s is positive (~+0.40) in today’s dashboard, aligning with the broader narrative that the classic curve inversion alarm is no longer active. Steepening can be “late-cycle,” but for next-90-days recession risk it’s typically less consistent with imminent contraction than a deeply inverted curve.

4. Credit spreads remain tight (HY OAS ~260 bps)
   High-yield spreads near ~260 bps indicate no meaningful default-cycle pricing. In most recession approaches, HY spreads widen decisively before the labor market prints undeniable weakness; we’re not seeing that yet.

5. Soft demand is weak (UMich sentiment: 55.2, final Aug 2026)
   Consumer sentiment at 55.2 (final August) is a clear “confidence tax” on discretionary demand. Weak sentiment doesn’t cause recessions by itself, but it often amplifies the impact of restrictive policy or labor softness once those appear. (sca.isr.umich.edu)

6. Policy risk is rising: Fed chair signals hikes may be needed if inflation stays stubborn
   The main regime shift over the past week is tone: Fed Chair Kevin Warsh has publicly opened the door to possible additional rate hikes if the Committee lacks confidence inflation is returning to 2% “clearly and at sufficient speed.” That raises the risk of a policy-driven confidence/financial-conditions shock, even while the economy’s hard data remains okay. (federalreserve.gov)

Category Breakdown

• Primary Indicators: 3 safe / 4 watch / 2 danger
  Mixed but stable: labor triggers are mostly safe, yet forward labor internals (temporary help) and certain “trend” measures keep the primary bucket from clearing.

• Secondary Indicators: 2 safe / 0 watch / 1 danger
  A modest negative tilt persists—enough to keep the score in MODERATE, not enough to escalate by itself.

• Housing & Construction: 0 safe / 1 watch / 1 danger
  Housing remains a meaningful drag: permits are watch and starts are warning—consistent with rate sensitivity and affordability pressure.

• Business Activity: 2 safe / 1 watch / 0 danger
  Business activity isn’t collapsing. This aligns with nowcasts that imply continued growth, even if uneven.

• Consumer Credit Stress: 1 safe / 2 watch / 1 danger
  The consumer is the swing factor: delinquencies and low savings imply reduced shock-absorption if unemployment rises.

• Market Signals: 7 safe / 2 watch / 5 danger
  A bifurcation: index levels/volatility look fine, but valuation and risk-ratio indicators remain elevated—more “late-cycle fragility” than “recession imminent.”

• Liquidity: 0 safe / 1 watch / 2 danger
  Liquidity is not comfortable: the ON RRP facility is near depleted and fiscal/interest expense metrics are flashing.

• Real-Time / High-Frequency: 0 safe / 1 watch / 1 danger
  High-frequency signals are split: jobless claims are fine, but freight/real-time cyclicals are weak.

Biggest Movers

1. ON RRP Facility ($7B): +764.5% (7D) — Contradictory / ambiguous
   A sharp jump from extremely low levels reflects short-term cash-management dynamics more than growth fundamentals. Directionally, a more “active” RRP can indicate tighter money-market plumbing, but given the low base, it’s best read as liquidity noise with potential signaling value, not a recession trigger.

2. Yield Curve (2s10s) (0.40): -30.8% (7D) — Confirmatory (slightly worsening)
   The curve remains positive, but the 7D move suggests some flattening. This is not a recession call by itself, but it does reduce the “all clear” benefit from steepening.

3. Yield Curve (2s30s) (0.88): -21.1% (7D) — Confirmatory (slightly worsening)
   Similar message: long-end relative moves can reflect term premium/inflation risk, but the direction is toward less steep.

4. Housing Starts (1239K): -19.7% (7D) — Confirmatory (worsening)
   Housing is one of the cleanest rate-sensitive cyclical channels. A ~20% drop (7D) reinforces that construction momentum is fragile.

5. VIX (14.4): -13.7% (7D) — Contradictory (improving)
   Falling volatility is usually “risk-on” and inconsistent with imminent recession, but it can also reflect complacency late-cycle—especially when policy rhetoric turns more hawkish.

90-Day Indicator Trends

No data available for this window.

(Note: the provided “90-day history” block contains date-stamped observations primarily in early-to-late June 2026 for many series, plus “today’s readings” for Sep 2, 2026. Where the history is sparse beyond June, we focus on the direction from the historical anchor to today.)

Labor: still strong on triggers, weaker on internals

• Initial claims: ~215k–229k in the June window, now 203k (week ending Aug 22). That’s a downshift, reinforcing the “no layoffs wave” story.
• Sahm Rule: 0.13 → 0.10 during June observations, now -0.03 (July) in today’s summary—improving further (less recession-like).
• Temporary help services: sits in DANGER and is described as a sharp decline (today: 2,505k). Historically, temp help is often an early labor-market crack—this is the cleanest “hard” recession-leading warning in the dashboard.

Interpretation: the recession playbook is currently inverted: headline labor stress is absent, but composition is deteriorating. That configuration fits a slow-growth late-cycle economy where firms hoard core labor while cutting flexible labor first.

Housing: deterioration is persistent

• Housing starts: June anchor around 1,465k then a drop to 1,177k in the June series, while today reads 1,239k (WARNING)—still below the earlier level and consistent with ongoing rate sensitivity.
• Building permits: 1,423k → 1,413k in June observations; today 1,433k (WATCH) suggests modest stabilization, but at a level that still implies slowing rather than re-acceleration.

Interpretation: housing is not in freefall, but it’s not providing cyclical lift. If the Fed tightens financial conditions further (explicitly or via guidance), housing is a likely transmission channel.

Financial conditions & credit: still not recessionary

• HY OAS: moved from roughly ~271–280 bps in early June observations to ~263–266 bps later in June; today ~260 bps—tight and broadly improving.
• NFCI: around -0.49 to -0.51 in June observations; today -0.57—even looser conditions.
• VIX: June window had spikes into the low 20s; today 14.4, signaling calmer markets.

Interpretation: credit is acting like the economy is fine. For the recession score to rise materially, you’d typically need to see spreads widen and/or financial conditions tighten consistently—not just for a day or two.

Growth nowcasts: expansion signal still intact

• Atlanta Fed GDPNow: 4.8% for 2026:Q3 as of Sep 1, up from 4.6% on Aug 26—hard to square with a near-term recession call. (atlantafed.org)
• NY Fed Staff Nowcast: 2.2% for 2026:Q3 (Aug 28)—slower than GDPNow but still expansionary. (newyorkfed.org)

Interpretation: nowcasts can be noisy, but the combined message is “not collapsing.”

Stock Screener Signals

Today’s quant screen is dominated by value/dividend and oversold growth flags: ARCC, AIG, BBY, FNF, HMC, T, BCE, plus oversold growth names like CHTR and TLK. In macro terms, this mix usually indicates two simultaneous market beliefs: (1) a preference for cash-flow durability and yield (defensive tilt), and (2) opportunistic mean reversion in select beaten-down growth where the market is pricing in too much cyclical or rate risk.

What stands out is that several flags sit in financials/credit-adjacent areas (AIG, ARCC, FNF) and consumer cyclicals (BBY). That’s not a “panic for safety” list (no heavy clustering in utilities/staples-only), and it’s not a “pure risk-on momentum” list either. It’s more consistent with a late-cycle barbell: investors want income + value as a hedge against policy and growth uncertainty, while selectively bottom-fishing in rate-sensitive/levered names that have already been de-rated.

One caution: the displayed dividend yields (e.g., “1002%”) are almost certainly data-quality artifacts (special distributions, stale prices, or parsing issues). Treat the style flags (value dividend / oversold growth) and RSI as the main signal, not the absolute yield print.

Latest Economic Developments

Fed messaging has turned meaningfully more hawkish into September. In his Jackson Hole remarks on Aug 28, 2026, Fed Chair Kevin Warsh indicated policymakers may need to do more if inflation is not clearly returning toward 2% at a sufficient pace—language markets interpreted as reopening the possibility of hikes. (federalreserve.gov) This matters for recession risk not because rates are high in an absolute sense, but because expectations and financial conditions can tighten quickly when the Fed shifts the reaction function.

Labor market data remains recession-inconsistent. Initial claims fell to 203,000 (week ended Aug 22, reported Aug 27), still near historic lows—evidence layoffs remain limited. (apnews.com) As long as claims stay in the low-200s, it’s difficult for near-term recession odds to surge unless a separate credit shock emerges.

Consumer confidence is weak and deteriorating at the margin. The University of Michigan’s final August 2026 sentiment reading shows a weak backdrop at 55.2, down versus July, consistent with households feeling pressure even if employment is stable. (sca.isr.umich.edu)

Markets are showing rate sensitivity rather than recession stress. U.S. equities pulled back on Sep 1 as Treasury yields moved higher (10-year referenced around 4.79%), with pressure concentrated in big-tech leadership. (apnews.com) This is a financial-conditions story—important because sustained yield rises can weigh on housing and discretionary spending even absent labor weakness.

Growth nowcasts remain firmly expansionary. Atlanta Fed GDPNow sits at 4.8% for 2026:Q3 as of Sep 1 (up from Aug 26), while the NY Fed Staff Nowcast is 2.2% for 2026:Q3 (Aug 28). (atlantafed.org) The spread between them underscores uncertainty, but neither is a recession print.

Near-Term Outlook (Next 30 Days)

The next 30 days are set up as a policy-and-labor confirmation window. The score is most likely to remain in the low-to-mid 30s unless one of two things happens: (1) weekly claims begin a clear upshift over multiple prints, or (2) Fed communication tightens financial conditions enough to hit hiring and discretionary demand.

Key catalysts:

• FOMC decision (Sep 16, 2026): With hawkish tone elevated post–Jackson Hole, guidance could shift markets even without an immediate hike. The Atlanta Fed’s own data page has shown elevated market-implied hike probability into Sep 16 in late August updates, reinforcing that this meeting is “live” for conditions. (atlantafed.org)
• Inflation releases ahead of the meeting: Any upside surprise that validates the Fed’s hawkish rhetoric could push yields and the dollar higher, tightening conditions.
• Labor data (payrolls/unemployment): The unemployment rate is already in WATCH territory at 4.1%. A move higher combined with rising claims would be the fastest path to a higher risk score.

Base case for the next month: moderate expansion with volatility around rates, not a recession impulse—unless the Fed catalyzes a sharper tightening in conditions.

Long-Term Outlook (3-6 Months)

Three forces will determine whether “moderate risk” drifts down toward low risk or up toward elevated:

1. Labor rebalancing vs labor break:
   If the economy continues to cool via lower quits and slower hiring while layoffs remain contained, recession risk can stay moderate and eventually fall. If temp help weakness spreads into broader layoffs, then the Sahm Rule can move quickly—at which point the score would re-rate materially.

2. Policy path and credibility:
   Warsh’s communication has re-centered inflation risk. (apnews.com) Over 3–6 months, the recession question becomes: does the Fed manage a “confidence cooling” without triggering a labor unwind? That’s historically difficult late-cycle, especially with low household savings (today: 3.0%) and rising consumer credit stress.

3. Cyclicals vs credit:
   Freight and goods-cycle weakness can persist without a recession if services remain firm. But if credit spreads begin to widen decisively from today’s tight levels, the probability of a broader downturn rises quickly.

Net: the next 3–6 months look like a late-cycle balancing act. The data today do not support a recession as the base case, but the distribution is fatter-tailed than the headline score alone suggests because policy error and confidence shocks can propagate fast when buffers (savings, fiscal flexibility) are limited.

What to Watch

Hard triggers (score-up conditions):

• Initial jobless claims: watch for a sustained move above ~230k–250k across several weeks (not a one-off).
• Sahm Rule: any climb toward 0.30+ would be an early warning; 0.50 is the formal trigger.
• HY OAS: a decisive widening (e.g., sustained move >350–400 bps) would be a classic “risk-off / recession pricing” confirmation.

Rate/financial-conditions catalysts:

• Sep 16, 2026 FOMC: hike or hawkish guidance that pushes long yields higher.
• 10-year yield behavior: continued upward pressure (as seen into Sep 1 trading) can hit housing and risk assets. (apnews.com)

Cyclicals / real economy internals:

• Temporary help: stabilization vs continued declines (often an early labor-market tell).
• Freight: any rebound would be an “all-clear” signal for goods-cycle stabilization; continued weakness keeps risk pinned in MODERATE.

Sources

No data available for this window.