Recession Risk 38/100 — July 29, 2026

Recession Risk Score: 38/100 — MODERATE (+1 vs 30 days ago)

Today’s Recession Risk Score is 38/100 (MODERATE), up +1 point from 30 days ago (37 on June 29, 2026). The macro picture still looks like a decelerating expansion, not an imminent contraction: the labor market’s layoff signal remains exceptionally tight, and broad financial stress is absent. The score’s upward drift is being driven less by “hard” recession mechanics (claims, spreads, unemployment shock) and more by late-cycle fragility (temp help, depressed sentiment, goods-cycle stress proxies) that tends to matter if—and only if—it bleeds into layoffs.

Score Trend — Last 30 Days

Over the last 30 days (June 29 → July 29, 2026), the score moved from 37 to 38 (+1), with a min of 33, a max of 38, and a 30-day average of 36. The distribution is notable: the system spent meaningful time in the mid-30s, but repeatedly snapped back to the high-30s ceiling.

The shape of the last 10 readings looks like choppy mean-reversion with a higher ceiling: repeated dips to 34 (e.g., July 19, July 22–23, July 28) were followed by quick rebounds to 38 (e.g., July 20–21, July 24, July 26–27, July 29). In practical terms, the economy/market complex is not “breaking,” but it is sensitive to marginal bad news—especially anything that would turn hiring caution into layoffs.

Key Drivers

1. Layoffs remain exceptionally contained (strongest near-term recession “anti-signal”).
   Initial jobless claims fell to 187,000 for the week ending July 18, 2026, the lowest since 1969 per contemporaneous reporting, and DOL confirmed the same headline level. (apnews.com)
   Implication: As long as claims remain sub-220k-ish with no sustained upshift, the probability of a recession starting in the next ~90 days typically stays capped.

2. Sahm Rule remains far from triggering (unemployment rise is modest, not shock-like).
   Your reading sits at 0.07 (SAFE), with the 90-day history showing the indicator drifting down from 0.20 (Apr 30) to 0.13 (mid/late May) and now lower. Implication: The labor market may be cooling, but it is not (yet) displaying the type of unemployment acceleration that usually marks recession onset.

3. Growth is slowing but remains positive in nowcasts (deceleration, not stall).
   Atlanta Fed GDPNow’s estimate for 2026:Q2 was ~1.7% SAAR around mid-July (e.g., 1.7% on July 16 in Atlanta Fed commentary; their research dashboard also showed 1.7% updated July 17). (atlantafed.org)
   Implication: A positive nowcast doesn’t eliminate recession risk, but it shifts the base case toward continued expansion unless labor/credit turns.

4. Soft data and labor-market “early cracks” are flashing late-cycle caution.

   • Temporary Help Services: 2,499k (DANGER) — historically a sensitive pre-layoff margin.
   • JOLTS quits rate: 1.9% (WARNING) — consistent with reduced worker confidence and lower job-switching power.
     Implication: These are the kind of indicators that often worsen before headline payrolls deteriorate.

5. Manufacturing is expansionary—but employment remains weak (goods cycle vs jobs cycle divergence).
   ISM Manufacturing PMI for June 2026 was 53.3 (sixth straight month of expansion), but ISM commentary highlights employment has been stuck in contraction territory for most of the last several years and remained below 50 in June. (ismworld.org)
   Implication: Output can hold up even as labor demand deteriorates at the margin—this is a classic “late-cycle productivity/throughput” look that becomes dangerous if demand cools further.

6. Financial conditions remain supportive; credit spreads are tight (no funding-stress transmission).
   Your HY OAS is ~281 bps (SAFE) and NFCI is -0.55 (SAFE)—a configuration that is historically inconsistent with imminent recession dynamics unless it flips quickly. Implication: For the next 90 days, credit is the fuse: if spreads stay tight, the economy usually needs a labor shock to tip.

Category Breakdown

• Primary Indicators: 3 safe / 4 watch / 2 danger
  Mixed but not recessionary: the “core” system is being held together by strong real-time labor and benign recession-rule readings, while late-cycle labor composition (temp help) injects risk.

• Secondary Indicators: 2 safe / 0 watch / 1 danger
  Secondary signals aren’t confirming a break—risk is present, but not broadening.

• Housing & Construction: 0 safe / 1 watch / 1 danger
  Housing remains a drag, consistent with a higher-rate world and low affordability—important mainly as a growth headwind, not a near-term recession trigger unless employment weakens.

• Business Activity: 2 safe / 1 watch / 0 danger
  Business activity looks decelerating-but-stable, consistent with a soft landing path.

• Consumer Credit Stress: 0 safe / 3 watch / 1 danger
  Stress is creeping (delinquencies, low savings, debt service), raising vulnerability if layoffs rise.

• Market Signals: 7 safe / 2 watch / 5 danger
  Markets are near highs and volatility is contained, but valuation/GDP-style measures are stretched—this reads more like late-cycle risk pricing than recession pricing.

• Liquidity: 0 safe / 1 watch / 2 danger
  Liquidity is less forgiving (notably the near-depleted RRP), increasing sensitivity to shocks, even if not currently manifesting as credit stress.

• Real-Time / High-Frequency: 0 safe / 1 watch / 1 danger
  Real-time is split: claims are strong, but the warning cluster (temp help, freight/goods proxies) keeps the “now” data from being unambiguously green.

Biggest Movers

1. Bank Unrealized Losses ($5,155B): +931.1% (7D) — Confirmatory (worsening risk)
   Even if partly a data-step/level-reset effect, a jump of this magnitude is a reminder that duration exposure can re-emerge as a confidence/liquidity problem under stress.

2. NY Fed Recession Probability (5.9%): -82.5% (7D) — Contradictory (improving risk)
   A sharp drop is consistent with the broader message from claims/credit: recession odds are not rising mechanically.

3. ON RRP Facility ($1B): -74.1% (7D) — Confirmatory (worsening fragility)
   Less cash parked at the Fed can mean less immediate liquidity backstop in money markets if volatility spikes.

4. Yield Curve (2s10s) (0.35): -10.4% (7D) — Slightly confirmatory (worsening risk at the margin)
   The curve remains positive but moved toward less steepness; the bigger point is the post-inversion regime is often late-cycle.

5. VIX (18.7): -6.8% (7D) — Contradictory (improving risk / complacency)
   Lower implied volatility typically aligns with risk-on conditions; however, it can also indicate complacency rather than macro safety.

90-Day Indicator Trends

The key takeaway from the 90-day window provided is macro stability with pockets of late-cycle deterioration—especially in labor composition, sentiment, and goods-cycle proxies—while financial stress remains absent.

• Labor market (real-time): improving on layoffs, weakening on “confidence.”

  • Initial claims: In the historical window shown (May), claims oscillated around 189k–211k, and today sits at 187k, reinforcing a downshift in layoff pressure.
  • JOLTS quits: Drifted from ~2.0% in mid/late May to 1.9% (WARNING) today—small numerically, meaningful behaviorally: fewer quits = lower worker bargaining power.

• Sahm Rule: trending safer.
  Sahm moved from 0.20 (Apr 30) to 0.13 (mid/late May) and is 0.07 today. That is a material improvement over ~90 days, and it directly argues against an imminent recession call.

• Growth nowcast: steady, low-but-positive.
  GDPNow is basically flat around ~1.8% through the window you provided, consistent with a slow-growth regime rather than contraction.

• Housing: softening at the margin.

  • Housing starts: 1502k (Apr 30) → 1465k (May 22/23) → 1427k today (WATCH): a clear downtrend from the earlier spring level.
  • Building permits: bounced in late May (1372k → 1442k), yet today is 1374k (WARNING)—a reversal back toward softness.

• Credit conditions (market-based): still supportive.

  • HY OAS: mostly ranged high-200s/low-300s in May; today ~281 bps—still “tight.”
  • NFCI: steady around -0.52 in May; today -0.55, consistent with loose conditions.

• Consumer vulnerability: rising sensitivity to shocks.
  The personal savings rate in your history fell from 4.0% (Apr 30) to 3.6% (May onward), and your current reading is 3.0% (WARNING)—an additional deterioration in cushion. Meanwhile, credit card delinquency ~2.9% stayed elevated, implying households are less able to absorb a labor-market hit.

• Sentiment: still recession-like even after a bounce.
  Michigan sentiment improved from 44.8 in May to 49.5 in June, but remains extremely depressed by historical standards. (isr.umich.edu)
  Your current indicator table still flags Consumer Sentiment as DANGER, consistent with a “feel-bad economy” that can become real if jobs weaken.

Stock Screener Signals

Today’s screener is dominated by “value dividend” flags (ARCC, AIG, BBY, FNF, HMC, T, LTM, BCE) with a smaller pocket of “oversold growth” (CHTR, TLK). That mix suggests markets are simultaneously (1) reaching for carry and (2) selectively probing idiosyncratic oversold opportunities—more consistent with a late-cycle grind than a broad recession panic.

The “value dividend” cluster typically aligns with defensive positioning (cash flow, yield, balance-sheet optics) when investors think growth will slow but not collapse. However, the printed yields (e.g., ARCC “1002%”) are clearly data-quality artifacts rather than investable reality; the useful signal is the style tilt: carry + value + defensiveness beats “high-beta cyclicals” in what the quant lens sees as today’s environment.

The oversold growth names (e.g., Charter with RSI 28) point to selective mean-reversion rather than broad risk-off. In recession setups, you tend to see wider screens lit up (more deep oversold breadth, credit-linked equities breaking). Here, it’s narrower—consistent with a moderate recession score.

Latest Economic Developments

• Labor market headlines reinforced the “no layoffs” message.
  The Department of Labor reported initial claims at 187,000 for the week ending July 18, down 22,000 from the prior week, with the four-week average also falling (per reporting). (apnews.com)
  This is the single most important near-term macro datapoint for recession timing: layoffs remain subdued.

• GDPNow remained positive but modest for 2026:Q2.
  The Atlanta Fed’s GDPNow estimate was ~1.7% SAAR in mid-July (e.g., July 16), indicating growth is slower than trend but still expanding. (atlantafed.org)

• Fed communications stayed inflation-focused heading into the July 28–29 meeting window.
  Reporting highlighted continued emphasis on delivering price stability and warnings that inflation may not be returning sustainably to 2% yet, reflecting a policy stance that is cautious about premature easing. (apnews.com)
  That matters for recession risk because the main pathway from “slow growth” to “recession” in 2026 is likely policy staying restrictive long enough to fracture labor/credit—not today’s growth level itself.

Near-Term Outlook (Next 30 Days)

Base case for the next month: risk score oscillates in the mid-to-high 30s, with direction determined by whether we see any regime change in layoffs or credit.

Key catalysts (and why they matter):

• Weekly jobless claims (each Thursday): You’re looking for a sustained move in the 4-week average rather than a one-off print. The recession score will rise quickly if claims start printing persistently above the low-200k zone and trend higher.
• Next payrolls/unemployment release: The market doesn’t need a spike to 5%—it needs acceleration. A move that pushes the Sahm reading meaningfully upward would reprice the whole “soft landing” narrative.
• FOMC decision/communication: The critical risk is “higher for longer” rhetoric if inflation progress stalls, which can tighten financial conditions even before spreads blow out.

Long-Term Outlook (3-6 Months)

Over a 3–6 month horizon, the economy looks more vulnerable than the 90-day recession score implies, but the vulnerability is conditional. The conditional chain is:

1. Soft data + temp help weakness persists →
2. hiring slows further →
3. layoffs begin to rise →
4. consumer cash-flow stress (low savings, rising delinquencies) amplifies demand weakness →
5. credit spreads widen → recession mechanics turn “on.”

Right now, we are clearly in steps (1)–(2), flirting with (3) via composition measures, but the hard confirmation (claims, spreads) is not here. The 90-day trajectory supports a “slow growth” regime: GDP nowcasts remain positive, financial conditions are loose, and recession rules (Sahm-like) are not close to trigger. The most important longer-term risk embedded in your dashboard is fragility: low savings + higher debt service + high valuations can turn a modest shock into a sharper downturn.

What to Watch

Labor (highest priority)

• Initial claims: Watch for a sustained climb in the 4-week average; a persistent upshift is your earliest reliable “recession clock” in real time.
• Unemployment rate: Track any acceleration that lifts the Sahm Rule meaningfully above today’s 0.07.

Credit / financial conditions

• HY OAS: If spreads break out from the high-200s into a persistent widening regime, the odds of a recession within subsequent quarters rise sharply.
• Bank unrealized losses + liquidity plumbing: Large step-ups (even if partly measurement) plus depleted RRP increases tail sensitivity if volatility rises.

Housing

• Permits/starts: Permits staying weak and starts trending down would reinforce a slow-growth backdrop and raise downside risk if labor cools.

Goods-cycle proxies

• Freight / copper-to-gold style signals: These are already warning; the key is whether the weakness spreads into employment/payrolls rather than staying contained in goods.

Sources

No data available for this window.