Recession Risk 34/100 — September 21, 2026

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

Today’s Recession Risk Score is 34/100 (MODERATE), and the direction of travel has been down over the past month (-4 vs 30 days ago). The macro message is “late-cycle but not breaking”: the highest-frequency labor stress triggers remain inactive, while financial conditions are still easy. The main recession-adjacent warning is concentration—weakness is showing up in interest-sensitive housing and early-cycle labor (temp help) rather than in broad layoffs or widening credit spreads.

Score Trend — Last 30 Days

The last 30 days show a controlled drift lower in recession risk: Start 38 → End 34 (Δ -4), with a range-bound profile (Min 34, Max 38, Avg 36; 31 samples). That’s important: when recession risk truly accelerates, the score typically doesn’t oscillate cleanly between two levels—it ratchets upward as multiple categories flip from “watch” to “warning/danger” together.

The shape is mean-reverting rather than trending: several “risk-on” conditions (tight spreads, low equity vol, easy NFCI) have capped downside tail-risk pricing, while the real economy’s soft patch (housing, sentiment, temp help, freight) has prevented a full downgrade into “LOW.” In the last 10 readings, the pattern alternates between 34 and 38, implying a market and macro system that is absorbing shocks without a cascade—for now.

Key Drivers

1. Labor stress signals remain OFF (high-frequency)

   • Initial Jobless Claims: 196k for the week ending Sep 12, 2026—a notably low print that keeps the near-term recession “tripwires” quiet. (apnews.com)
   • Sahm Rule: -0.07 (SAFE) (well below trigger) — consistent with an expansion regime in the next ~90 days.

2. The Fed just tightened—policy headwind rises into Q4/Q1

   • On Sep 16, 2026, the FOMC raised the target range 25 bps to 3.75%–4.00%. (federalreserve.gov)
   • This doesn’t usually hit growth instantly; it typically shows up with a lag via housing/credit creation and business confidence.

3. Growth tracking is not near-stall—GDPNow is strong

   • Atlanta Fed GDPNow estimates 2026:Q3 real GDP at ~5.1% SAAR as of Sep 17, 2026. (atlantafed.org)
   • That is not the profile of an economy “rolling over” into imminent recession—though nowcasts can swing with each data release.

4. Financial conditions remain loose, not recessionary

   • Chicago Fed NFCI: about -0.56 (Sep 11) — clearly on the “easy” side of neutral. (fred.stlouisfed.org)
   • Easy conditions dampen the probability of a fast, liquidity-driven downturn.

5. Credit spreads are tight (no stress signal)

   • HY OAS ~270 bps (Sep 17) — tight enough to argue against imminent recession or a credit event. (convextrade.com)

6. Housing is weakening (rate sensitivity is biting)

   • August 2026 housing starts: 1.275M SAAR, and permits: 1.394M SAAR—both pointing to continued softness in interest-sensitive activity. (census.gov)

Category Breakdown

(Using your CATEGORY BREAKDOWN counts.)

• Primary Indicators: 3 safe / 4 watch / 2 danger
  Mixed but stable: labor stress is mostly contained, yet temp help is a genuine leading red flag and the unemployment rate is drifting higher.

• Secondary Indicators: 2 safe / 0 watch / 1 danger
  Thin but slightly favorable: not many “yellow lights,” but the one danger reading here matters because second-order indicators tend to turn after the first cracks.

• Housing & Construction: 0 safe / 0 watch / 2 danger
  Cleanest weak spot in the dashboard: housing remains a persistent drag and the clearest transmission channel for tighter policy.

• Business Activity: 2 safe / 1 watch / 0 danger
  Business-side signals are not collapsing; this supports the base case of slower growth, not contraction.

• Consumer Credit Stress: 1 safe / 2 watch / 1 danger
  Stress is creeping, not spiking—watch delinquencies + savings cushion closely because this category can flip quickly if the labor market loosens.

• Market Signals: 6 safe / 3 watch / 5 danger
  Markets are sending a split message: low vol and strong index levels vs. valuation/fiscal/ratio-based “danger” flags. That’s “late-cycle risk,” not a near-term recession trigger by itself.

• Liquidity: 0 safe / 1 watch / 2 danger
  Liquidity indicators are flashing caution—important mainly because liquidity can convert a soft patch into a sharper drawdown if something breaks.

• Real-Time / High-Frequency: 0 safe / 1 watch / 1 danger
  The real-time set says “watch carefully”: claims are fine, but the danger flag (freight) suggests the goods economy is still under pressure.

Biggest Movers

Top 5 by absolute 7‑day % move (from your BIGGEST MOVERS block):

1. ON RRP Facility ($5B): -74.9% (7D)
   Confirmatory (worsening liquidity risk). Less RRP usage often reflects cash moving elsewhere; in late-cycle phases this can be benign, but it also reduces a “buffer” that previously absorbed reserves/cash fluctuations.

2. Yield Curve (2s30s) (0.81): +11.2% (7D)
   Contradictory (improving vs recession risk). Steepening typically argues against imminent recession; however, re-steepening after inversion can also be “late-cycle normalization,” so it’s supportive but not decisive.

3. NY Fed Recession Probability (0.9%): -10.6% (7D)
   Contradictory (improving). Lower model-implied probability reduces near-term recession odds. (Note: your “0.9%” label appears inconsistent with the “~21%” narrative—treat the direction as the actionable piece.)

4. VIX (15.4): -4.5% (7D)
   Contradictory (improving). Lower vol reflects calmer equity hedging demand and typically corresponds to looser financial conditions.

5. Chicago Fed NFCI (-0.56): -2.2% (7D)
   Contradictory (improving). More negative NFCI implies conditions easing further, which historically delays recession dynamics.

90-Day Indicator Trends

Your 90‑day history snapshot is incomplete (many series stop in mid‑July), but there’s still enough to identify direction-of-travel and whether the macro mix is “tightening” or “loosening.”

Labor & income: stable stress gauges, but early-cycle cracks persist

• Initial claims moved from 226k (Jun 23) to 215k (late Jun/early Jul prints) in the provided history and sits at 196k (week ending Sep 12) today—improving meaningfully into September. (fred.stlouisfed.org)
• Sahm rule drifted from 0.10 (Jun 23) to 0.07 (early Jul) and is -0.07 today: still expansion-consistent and moving away from trigger.
• Temporary help remains the biggest labor outlier: the history shows it around 2.49–2.50M in late June/early July, while today you flag 2.52M (DANGER) with a “sharp decline” characterization. Regardless of the exact level, the signal is key: temp help is a classic “first to turn” series and is consistent with late-cycle cooling.

Consumer: sentiment is weak; savings cushion is thin

• UMich sentiment improved from deep lows in late June/early July in your history but remains weak today (55.2, WARNING)—a reminder that “better” sentiment can still be “bad” sentiment.
• Personal savings rate rose from 2.6% (Jun 23) to 3.0% (late Jun onward)—an improvement but still a very low cushion in your framework.

Financial conditions: supportive (for now)

• NFCI hovered around -0.50 to -0.52 through late June/early July and is -0.56 as of Sep 11—conditions are easier, not tighter. (fred.stlouisfed.org)
• HY OAS oscillated roughly 265–283 bps in late June/early July history and is ~270 bps in mid‑September—still in a “carry regime,” not repricing stress. (convextrade.com)
• VIX trended down from the high‑teens toward the mid‑teens in the provided history, consistent with today’s 15.4.

Housing: consistently weak (and likely the main transmission channel of tighter policy)

• Your history shows housing starts around 1.177M (late Jun/early Jul) and today’s official August print is 1.275M with permits 1.394M—better than early-summer levels in the history but still below trend in your regime labels. (census.gov)
• The practical macro takeaway: housing is not collapsing, but it is not re-accelerating enough to offset policy drag.

Growth: nowcast strong, but the score keeps a “soft patch” discount

• The big divergence is between “hard” macro nowcasting (GDPNow ~5.1% SAAR) and several cyclical micro signals (freight, temp help, sentiment). That’s exactly the combination that produces a MODERATE score rather than LOW. (atlantafed.org)

Stock Screener Signals

Today’s quant list is dominated by value/dividend flags (ARCC, AIG, BBY, FNF, HMC, T, LTM, BCE) with a couple oversold growth names (CHTR, TLK). The market implication is not “panic”—it’s selective defensiveness: investors are willing to own cyclicality selectively, but they still prefer cash-flow visibility and lower multiples.

Two interpretations matter for recession risk:

1. Defensive carry + value tilt suggests late-cycle positioning, not a recession stampede.
   The low P/E clustering (many single-digit to low-teens) aligns with an environment where investors want margin of safety and income. That matches the macro mix: not recessionary labor stress, but enough uncertainty (policy, housing, confidence) to favor “boring” cash generators.

2. Oversold growth flags look more like mean-reversion trades than a broad de-risking event.
   Names like CHTR (RSI 28) and TLK (RSI 30) being highlighted alongside value dividends suggests markets are not uniformly pricing downside—they’re rotating within equities rather than exiting risk altogether. That’s consistent with tight HY spreads and low VIX.

One caveat: the listed dividend yields (some appear extremely high) look data-quality distorted—use the factor direction (value/dividend bias, oversold pockets) rather than the literal yield numbers as the signal.

Latest Economic Developments

• Federal Reserve: The Fed raised rates 25 bps on Sep 16, 2026 to a 3.75%–4.00% target range. The official statement frames the hike as supporting a timelier return to 2% inflation. (federalreserve.gov)
• Labor market (high-frequency): Initial jobless claims fell to 196,000 for the week ending Sep 12, 2026, the lowest since mid‑July—evidence that layoffs remain contained. (apnews.com)
• Growth tracking: Atlanta Fed GDPNow holds 2026:Q3 at ~5.1% SAAR as of Sep 17—a strong near-term growth signal. (atlantafed.org)
• Leading indicators: The Conference Board LEI edged down -0.1% m/m in August 2026 after +0.2% in July—softening, but not a collapse. (conference-board.org)
• Housing: The Census Bureau reported August 2026 starts 1.275M and permits 1.394M (SAAR), consistent with continued rate sensitivity. (census.gov)
• Financial conditions: Chicago Fed NFCI around -0.56 (Sep 11)—still signaling easy conditions rather than tightening stress. (fred.stlouisfed.org)

Near-Term Outlook (Next 30 Days)

Base case for the next month: continued expansion with rising tail risk, i.e., “soft patch” without broad labor break.

What can move the score meaningfully by late October 2026:

• Labor turn confirmation (fastest swing factor):

  • If claims’ 4‑week average breaks higher and sustains above your ~230k threshold, risk can reprice quickly.
  • Watch for diffusion: weakness moving from temp help into broad payroll deceleration.

• ISM / business surveys (cycle confirmation):

  • The next ISM manufacturing and services prints will matter mainly for new orders and employment components—the parts that lead hard data.

• Housing follow-through:

  • Building permits and starts can bounce month-to-month, but trend matters. A further downshift would reinforce policy drag and raise recession odds into 2027.

• Market-based tightening:

  • The score stays moderate as long as HY OAS stays tight and NFCI remains negative. A widening of spreads into a new regime (e.g., sustained move above ~350–400 bps) would be a different message than today’s 270 bps.

Long-Term Outlook (3-6 Months)

Over the next 3–6 months, the economy looks set up for a late-cycle test rather than a guaranteed downturn:

• Structural positives:

  • Labor stress indicators are not flashing.
  • Financial conditions remain easy, which tends to lengthen expansions.
  • GDPNow-strength implies momentum entering the quarter, giving a buffer against shocks.

• Structural drags:

  • Policy is tighter at the margin after the Sep 16 hike, and housing remains the cleanest transmission channel.
  • Consumer resilience is thinner with a low savings rate and creeping credit stress.
  • Cyclical “micro” deterioration (temp help + freight) is consistent with late-cycle slowdowns that sometimes precede broader labor weakening.

Historical parallel (pattern, not prediction): late-cycle soft patches often look like (1) housing weak → (2) temp help weak → (3) quits decline → (4) claims rise. You are currently between steps (1)–(3). The question for Q4 2026 and early 2027 is whether the sequence progresses to (4) or stalls out.

What to Watch

Thresholds and events most likely to move the score:

• Initial jobless claims (weekly):
  • Key threshold: 4‑week MA > ~230k (your framework) and rising for multiple weeks.
• Temp help services (monthly):
  • Any continued deterioration is a high-signal early warning; stabilization would reduce tail risk materially.
• HY OAS / credit stress (daily/weekly):
  • Watch for regime shift: sustained spread widening rather than a 1–2 day move.
• Chicago Fed NFCI (weekly):
  • A move toward 0 would imply tightening conditions; staying around -0.5 remains supportive.
• Housing (monthly):
  • Permits and starts trend; a sustained slide would reinforce recession risk into 2027.
• Conference Board LEI (monthly):
  • If LEI turns from small declines to persistent contraction, recession probability rises with a lag.

Sources

No data available for this window.