Recession Risk 34/100 — September 1, 2026

Recession Risk Score: 34/100 — MODERATE (-4 vs 30 days ago)

Today’s Recession Risk Score is 34/100 (MODERATE), down 4 points from 30 days ago (38 → 34). The signal mix still argues for a soft-but-still-growing baseline over the next 90 days: labor-market “break” indicators remain untriggered, and credit/financial conditions are not tightening meaningfully. The main offset is a fragile household backdrop (weak confidence, low savings, creeping delinquencies) that raises downside convexity if joblessness starts to rise. Policy risk remains two-sided, with recent Fed communication and market pricing leaning more hawkish into mid-September.

Score Trend — Last 30 Days

The last 30 days show a gentle mean-reversion lower in recession risk: Start 38 → End 34 (Δ -4), with a range of 34–42 and an average near 37. The path wasn’t a straight glide—risk spiked toward the upper bound earlier in the window, then drifted back toward the 34 floor that has repeatedly held.

The last 10 readings show a choppy “two-step” pattern: several prints at 38 interspersed with quick drops back to 34. That shape is consistent with a market and data environment where headline-driven shocks (policy rhetoric, geopolitics, inflation prints) can briefly elevate risk, but the core high-weight recession triggers (labor deterioration, credit spread blowout) are not confirming—so the score snaps back toward its mean.

Key Drivers

1. Labor triggers remain OFF (high weight, high confidence)

   • Sahm Rule: -0.03 (SAFE) — clearly untriggered, keeping the top-weight near-term recession alarm dormant.
   • Initial claims: 203K (SAFE) — still consistent with low layoffs; the week ending Aug 22, 2026 printed 203,000 on FRED’s weekly claims tables, matching the widely covered release. (fred.stlouisfed.org)
   • Interpretation: As long as claims stay near ~200k and the Sahm metric remains sub-threshold, the model’s base rate of imminent recession remains capped.

2. Yield curve normalization reduces “classic” recession signal strength

   • 2s10s: +0.41 (WATCH) — positive, but moving enough to remain “watch” rather than “safe.”
   • 2s30s: +0.88 (SAFE) — more convincingly normal.
   • Interpretation: A positive curve doesn’t immunize the economy, but it removes one of the most historically reliable medium-lag warnings (persistent inversion).

3. Leading indicators improved at the margin

   • The Conference Board’s LEI rose +0.2% m/m in July 2026 to 99.5, following a small (revised) dip in June. (conference-board.org)
   • Interpretation: That’s not a boom signal, but it’s inconsistent with a broad-based, imminent contraction.

4. Financial conditions and credit stress remain benign

   • HY OAS: 260 bps (SAFE) — tight, not recessionary.
   • Chicago Fed NFCI: -0.57 (SAFE) — loose conditions.
   • Interpretation: Recessions that arrive “fast” usually show credit deterioration first. We do not have it.

5. Household fragility is the biggest downside amplifier

   • UMich sentiment: 55.2 (WARNING) — weak confidence.
   • Personal savings rate: 3.0% (WARNING) — thin buffer.
   • Credit card delinquency: 2.9% (WATCH) — elevated and a meaningful “slow burn” stress signal.
   • Interpretation: The economy can keep growing with weak sentiment—until it can’t. Low savings + rising delinquency means the consumer is less resilient if hiring softens.

6. Fed reaction-function risk is rising again

   • Recent Fed messaging has turned more openly hawkish. Fed Chair Kevin Warsh signaled that rate increases may be needed if inflation doesn’t fall, a theme reinforced by Jackson Hole coverage in the past several days. (apnews.com)
   • Atlanta Fed’s research dashboard also shows the market-implied probability of a hike by 2026-09-16 elevated (as of the latest dashboard reading). (atlantafed.org)
   • Interpretation: The dominant tail risk is no longer “credit accident now,” but policy over-tightening into a consumer that already looks stretched.

Category Breakdown

• Primary Indicators: 3 safe / 5 watch / 1 danger
  Labor is mostly holding (claims, Sahm), but unemployment, quits, and manufacturing labor sit in “watch,” consistent with cooling—not cracking.

• Secondary Indicators: 2 safe / 0 watch / 1 danger
  Secondary is broadly fine, but the “danger” count here reinforces that some slower-moving fundamentals remain soft even if the top-line economy keeps expanding.

• Housing & Construction: 0 safe / 1 watch / 1 danger
  Housing is the weakest cyclical pocket: permits are only moderate while starts are flagged as weak—still a headwind to broad activity.

• Business Activity: 2 safe / 1 watch / 0 danger
  Business-side data remain non-recessionary. This supports the idea of rotation (goods softness, services/AI capex resilience) rather than collapse.

• Consumer Credit Stress: 1 safe / 2 watch / 1 danger
  This is the “quiet risk” bucket: delinquencies and debt service are watch-level, and the low savings cushion is a late-cycle vulnerability.

• Market Signals: 7 safe / 2 watch / 5 danger
  Risk assets are strong (indices near highs, low VIX), but several valuation/ratio-based indicators are in danger—suggesting pricing is optimistic even as macro is mixed.

• Liquidity: 0 safe / 1 watch / 2 danger
  Liquidity is a yellow/red area, especially with the ON RRP near depletion, reducing a prior shock absorber.

• Real-Time / High-Frequency: 0 safe / 1 watch / 1 danger
  High-frequency is split: layoffs are fine, but freight/goods-flow signals are deteriorating.

Biggest Movers

1. ON RRP Facility ($7B): -44.7% (7D)

   • Confirmatory (worsening risk): Less liquidity parked in RRP can mean less spare cash buffer in the plumbing. It’s not automatically bad, but with other fragilities (bank unrealized losses), it’s not a comfort signal.

2. Yield Curve (2s10s): -30.8% (7D)

   • Mixed/confirmatory: It’s still positive, but the steepening impulse has cooled. Less “tailwind” from curve normalization means less improvement in classic recession odds.

3. NY Fed Recession Probability: +23.0% (7D)

   • Confirmatory (worsening risk): Direction matters more than level. Even if today’s level is low, a sharp 7-day rise is a reminder that term-structure-based recession models are sensitive to rates volatility.

4. Yield Curve (2s30s): -21.1% (7D)

   • Contradictory-to-risk if still positive, but watch the direction: A smaller spread is not inherently recessionary, but it suggests the curve is less decisively normalizing.

5. Housing Starts: -19.7% (7D)

   • Confirmatory (worsening risk): Housing is a classic leading sector; rapid downside moves often foreshadow broader cyclical cooling.

90-Day Indicator Trends

Important limitation: the “90-day history” provided in today’s data block contains partial windows for many series (often June-only snapshots). Where we have a short run of daily values, we can still identify direction of travel. Where we don’t, we’ll focus on what is observable and avoid inventing intermediate points.

Labor market: cooling but not breaking

• Initial claims (June sample shown): 215K–229K range early June; latest provided today is 203K (SAFE). That’s directionally better than early-June prints and inconsistent with a layoffs-led downturn.
• Sahm Rule (June sample): downshifted from 0.13 → 0.10 by mid-June; today is -0.03 (SAFE). Direction: improving, moving farther from trigger.
• Unemployment rate (history sample shows 4.3% in June entries; today reads 4.1% WATCH): direction suggests stabilization/lower, though it remains “watch” because any renewed rise can accelerate Sahm.

Credit & financial conditions: supportive

• HY OAS (June sample): moved from ~320 bps → ~263 bps by late June; today 260 bps. Direction: tightening spreads (improving), not the widening you’d expect pre-recession.
• NFCI (June sample): improved from around -0.49 to ~-0.51; today -0.57 (looser). Direction: more accommodative.

Goods/housing: the weak flank

• Freight index sits in DANGER today and throughout the provided sample—persistent weakness, not a one-off.
• Housing starts fell sharply in the June sample (1465K → 1177K in the mid-June change shown) and today remains weak at 1239K (WARNING). Direction: downtrend with only partial rebound—still a macro headwind.

Consumer stress: slow grind higher risk

• Savings rate in the sample is 2.6% (danger); today is 3.0% (warning)—a slight improvement but still low by historical standards.
• Credit-card delinquency is stuck around 2.9% in the sample and remains 2.9% today: not an acute acceleration in this window, but elevated enough to matter if labor softens.

Growth nowcasts and leading indicators: not collapsing

• Atlanta Fed’s dashboard shows GDPNow for 2026:Q3 at ~4.6% with an update timestamp in late August. (atlantafed.org)
• Conference Board LEI is positive on the latest monthly print (+0.2% in July). (conference-board.org)

Net: Over the “last 90 days” lens we can observe, the economy looks two-track: labor + credit are cushioning downside, while housing + goods flow + consumer resilience are the stress points that could transmit weakness if policy tightens.

Stock Screener Signals

Today’s screener is dominated by “value dividend” flags (ARCC, AIG, BBY, FNF, HMC, T, BCE, LTM) with a smaller set of “oversold growth” (CHTR, TLK). That combination typically appears when the market is not pricing imminent recession, but participants are still seeking carry and cash flows (dividends, defensives, balance-sheet value) rather than paying up for high-duration growth across the board.

Two notable interpretations:

1. Income-and-quality bias is still the market’s comfort trade. Names like insurers/financials and telecoms screening as value-dividend suggests investors prefer steady cash generation—a posture consistent with “moderate risk” rather than “risk-on boom.”
2. Selective oversold growth (e.g., CHTR with very low RSI) hints at idiosyncratic stress rather than systematic unwind. In recessions, oversold signals usually spread broadly. Here, the screen looks more like rotation and dispersion.

One caveat: the yields shown (e.g., triple-digit “yields”) are almost certainly data-quality artifacts (special dividends, trailing calculations, or corporate actions). The macro signal to take is the factor clustering—value/income leadership—not the raw yield magnitude.

Latest Economic Developments

• Fed: hawkish tilt re-emerges from Jackson Hole. In the last several days, reporting from Jackson Hole emphasized Fed Chair Kevin Warsh’s message that the Fed may need to raise rates if inflation remains stubborn. (apnews.com)
• Markets: rate-hike probabilities elevated into mid-September. The Atlanta Fed’s research dashboard lists a high market probability of a rate hike by 2026-09-16, underscoring that investors are treating the September meeting as live. (atlantafed.org)
• Labor: layoffs remain low in the latest weekly report. Weekly initial claims were widely covered as 203,000 for the most recent reported week, consistent with a still-stable labor market. (apnews.com)
• Leading indicators: modest improvement. The Conference Board reported the LEI up +0.2% in July 2026, with commentary pointing to continued expansion but consumer pressure from living costs—matching today’s “weak sentiment / low savings” tension. (conference-board.org)
• Energy/geopolitics: inflation tail risk. A fresh geopolitical flare-up has pushed oil prices higher in recent coverage, reinforcing the risk of upside inflation surprises that could provoke the Fed. (apnews.com)

Bottom line: the macro impulse from “hard data” (claims, LEI, growth tracking) is still non-recessionary, but the policy function is becoming a bigger driver of near-term risk.

Near-Term Outlook (Next 30 Days)

The next month is about whether we get confirmation of one of two regimes:

1. Soft-but-still-growing (base case):

   • Claims stay contained (roughly low-200k range) and unemployment does not accelerate.
   • Credit spreads remain tight and NFCI remains loose.
   • Fed stays on hold in September or communicates conditionality without tightening financial conditions sharply.

2. Policy-driven cooling shock (tail risk rising):

   • A hot inflation print or hawkish Fed communication lifts rate expectations and broad yields, tightening conditions into weak confidence and low savings.
   • Housing softens further and goods-side weakness bleeds into hiring.

Key known catalyst: the September 16, 2026 FOMC meeting (mid-September), with markets increasingly attentive to inflation-sensitive data beforehand. (apnews.com)

Long-Term Outlook (3-6 Months)

Over a 3–6 month horizon, the macro picture looks structurally split:

• Supports to expansion:

  • Labor is not sending recession alarms (Sahm untriggered; claims low).
  • Credit is not pricing distress (HY spreads tight; NFCI loose).
  • Growth tracking (GDPNow) is strong enough to argue against an imminent downturn if it holds. (atlantafed.org)

• Structural vulnerabilities:

  • Consumer resilience is thinning (weak sentiment, low savings, rising delinquencies).
  • Housing remains a persistent drag.
  • Liquidity buffers (RRP depletion) and bank unrealized losses elevate the nonlinear risk of a funding or confidence event if rates reprice higher.

Historical parallel (conceptual, not a one-to-one): late-cycle expansions often fail not because “growth is already negative,” but because policy and financial conditions tighten into a fragile household sector. Today’s data are not recessionary, but the distribution of outcomes is widening: moderate risk can flip quickly if the labor market turns.

What to Watch

Labor (highest priority)

• Initial claims 4-week average: watch for a sustained move up (not a one-week print).
• Unemployment rate: any climb that starts pulling Sahm meaningfully toward trigger is the fastest path to a higher score.

Credit & financial conditions

• HY OAS: a regime shift would look like persistent widening (e.g., a move from the ~260s into the 300s+ with follow-through).
• NFCI: a turn upward toward zero would signal tightening.

Housing

• Starts and permits: continued downside would reinforce a broader cyclical fade.

Fed / policy

• September 16, 2026 FOMC: watch for either a hike or guidance that re-prices the entire front end. (apnews.com)
• Energy-driven inflation impulse: oil spikes feeding headline inflation would raise the probability of a hawkish surprise. (apnews.com)

Sources

• No data available for this window.