Recession Risk 34/100 — November 18, 2026
US recession risk over the next 90 days is MODERATE (34/100): late-cycle softening signals are rising, but the core real-economy and labor-market data still argue against an imminent contraction. The Sahm Rule remains at 0.00 in October 2026 (no trigger), and weekly initial jobless claims are still very low at 200,000 for the week ending Nov 7, 2026. Financial conditions remain broadly supportive (Chicago Fed NFCI around -0.46 in your tracker) while high-yield spreads are only modestly elevated (ICE BofA US HY OAS ~3.34% as of Nov 16, 2026). The key tension is that forward/early-cycle warnings (weak housing permits/starts, temp-help deterioration, very poor sentiment, freight softness) are flashing amber/red even as GDP tracking remains positive (Atlanta Fed GDPNow for 2026 Q4 was ~3.5% as of Nov 9, 2026).
Recession Risk Score: 34/100 — MODERATE (-12 vs 30 days ago)
Today’s Recession Risk Score is 34/100 (MODERATE), and it has fallen by 12 points versus 30 days ago (46 → 34). The headline message is a familiar late-cycle split: coincident data (labor, profits, broad activity) remains sturdy, while forward-looking and early-cycle signals (housing pipeline, temp-help, freight, sentiment, select market internals) keep throwing off amber-to-red warnings. The score’s decline over the past month says the system is pricing less “imminent recession” risk, but it is not declaring “all clear” given the persistent leading softness.
Score Trend — Last 30 Days
Over the last 30 days (2026-10-19 → 2026-11-18), the risk score mean-reverted lower from a start of 46 to an end of 34 (-12), with a min of 33, max of 46, and average of 37. That profile is not a grind in one direction—it’s a decline with intermittent shock spikes, consistent with a market-and-data regime where recession odds are being repriced lower, but headline risk quickly flares on event-driven volatility.
The last 10 readings highlight this “calm punctuated by jolts” shape: the score dropped to 33 on 11/11, stabilized around 34 for several sessions, then jumped to 44 on 11/15, fell back to 34 on 11/16, spiked again to 42 on 11/17, and returned to 34 today. The implication: baseline conditions are moderate and improving, but the macro is highly sensitive to marginal changes in inflation expectations, AI-linked equity leadership, and any hint of labor-market rollover.
Key Drivers
1) Labor-market stress remains notably absent (keeps the score capped).
- Initial Jobless Claims: 200K (week ending Nov 7, 2026) remains “SAFE” in your dashboard.
- Sahm Rule: 0.00 (Oct 2026) remains “SAFE” and nowhere near a 0.50 recession trigger.
These are the two most important “don’t panic yet” inputs: they argue against an imminent contraction without a clear deterioration in unemployment dynamics.
2) Financial conditions are still supportive (but not a free lunch).
- Chicago Fed NFCI: ~-0.46 (“WATCH” in your tracker), which is still looser than average (negative is easier-than-average conditions). FRED/Chicago Fed representations around this level remain consistent with broadly accommodative conditions for risk assets and borrowers. (fred.stlouisfed.org)
- Credit is not in a stress regime: ICE BofA US HY OAS ~3.34% (Nov 16, 2026)—modestly elevated, not “blowout.” (equibles.com)
3) Growth impulse is positive, but the mix is uneven.
- Atlanta Fed GDPNow: Q4 2026 tracking ~3.4% as of Nov 17 (down slightly from ~3.5% on Nov 9). (atlantafed.org)
This matters: recession risk typically rises when nowcasts roll over and labor cracks. Right now, growth tracking is still expansionary, even if it is softening at the margin.
4) Housing pipeline remains a leading weak spot (keeps tail risk elevated).
- Your readings show Building Permits: 1,398K (WARNING) and Housing Starts: 1,221K (WARNING)—both below trend.
- The Census Bureau calendar confirms the Oct 2026 New Residential Construction release is scheduled for Nov 18, 2026, reinforcing that housing is a near-term catalyst category today. (census.gov)
Housing doesn’t have to collapse to matter: a sub-trend pipeline is enough to drag on interest-sensitive demand and upstream goods activity.
5) “Soft” leading indicators are flashing red: temp help, freight, sentiment.
- Temporary Help Services: 2,526K (DANGER) — historically a high-signal early warning when it declines meaningfully.
- Freight Transportation Index: -0.4 (DANGER) — the goods economy remains soft.
- UMich Sentiment: 48.1 (DANGER) — crisis-level pessimism even while payrolls remain positive.
This combination is a classic late-cycle configuration: hiring plans cool first (temps), goods slow next (freight), consumers feel bad early (sentiment), and only later (sometimes much later) does the headline labor market break.
6) Markets are near highs, but volatility and AI-linked concentration are making the regime jumpy.
- Major indexes have been swinging: AP reported the Dow fell ~1.6% and Nasdaq ~0.7% on Nov 17 amid AI-stock volatility. (apnews.com)
- Fed communications are increasingly focused on AI investment dynamics: NY Fed’s John Williams flagged AI-driven strength and the risk that persistent inflation could keep rates “higher for longer.” (axios.com)
Category Breakdown
Using your CATEGORY BREAKDOWN counts:
- Primary Indicators (3 safe / 5 watch / 1 danger): Mixed but not recessionary—core “primary” readings are mostly safe-to-watch, consistent with a moderate score rather than high risk.
- Secondary Indicators (2 safe / 0 watch / 1 danger): Secondary confirms late-cycle softening (not collapse), with one danger reading acting as a tail-risk anchor.
- Housing & Construction (0 safe / 0 watch / 2 danger): Housing is the clearest leading warning cluster right now—pipeline metrics are not stabilizing enough to downgrade risk.
- Business Activity (2 safe / 1 watch / 0 danger): Business activity is still holding up; this is a key reason risk is not higher today.
- Consumer Credit Stress (1 safe / 2 watch / 1 danger): Household stress is creeping higher (delinquencies/DSR), consistent with “late-cycle” rather than “early-cycle.”
- Market Signals (6 safe / 3 watch / 5 danger): Internals are bifurcated: headline indexes are safe, but several market-based macro ratios (and industrial-fear proxies like copper/gold) are red-flagging.
- Liquidity (0 safe / 1 watch / 2 danger): Liquidity is a vulnerability point—an area where “small” plumbing shifts can create outsized macro ripple effects.
- Real-Time / High-Frequency (0 safe / 1 watch / 1 danger): Real-time data remains conflicted—enough to keep the score in MODERATE, not LOW.
Biggest Movers
From your BIGGEST MOVERS block (top 5 by absolute 7-day % change):
-
ON RRP Facility ($182M): +678.5% (7D) — Contradictory / plumbing signal.
A large percentage move off a tiny base is not the same as system stress, but it reinforces that liquidity regime is shifting and should be monitored alongside bank funding and reserve dynamics. -
Yield Curve (2s10s) (0.29): -24.2% (7D) — Confirmatory (worsening tail risk) even though curve is positive.
A less-steep curve week-over-week can re-tighten marginal financial impulse. You’re still above zero, but the direction over the week leans risk-up at the margin. -
VIX (19.3): -13.5% (7D) — Contradictory (improving).
Lower implied volatility typically corresponds to easier financial conditions and reduced near-term recession pricing. -
Yield Curve (2s30s) (0.59): -11.0% (7D) — Mildly confirmatory (worsening).
The long-end steepness easing can be read as slightly less “growth optimism,” though it’s not recessionary by itself. -
SOS Recession Indicator (1.10): -8.3% (7D) — Contradictory (improving).
Lower insured unemployment stress is consistent with the “labor still fine” narrative.
90-Day Indicator Trends
Your provided 90-day histories mostly cover 2026-08-20 through 2026-09-21, which is a partial window, but it still reveals the direction of travel across several high-signal series:
-
Yield curve (2s10s): 0.46 → 0.25 (Aug 20 to Sep 21)
That is a clear flattening over the sample (late-cycle dynamic). Today’s reading (~0.29) remains positive, but the broader theme is: the curve is no longer “recession-imminent inverted,” yet it is also not accelerating into a fresh early-cycle steepening. -
Initial jobless claims: 209K → 196K (Aug 20 to Sep 21)
Claims trended down into mid-September—consistent with today’s “SAFE 200K” characterization. The key risk is not the level; it’s a persistent uptrend—which you do not have yet. -
Building permits: 1,443K → 1,394K (Aug 20 to Sep 21)
Permits drifted lower and then stepped down into “warning.” That’s a forward housing pipeline deterioration—often one of the earliest macro cracks. -
Housing starts: ~1,239K steady → 1,275K late (Aug 20 to Sep 20)
Starts were weak but not collapsing in your history. Today’s 1,221K reading suggests the category is still below trend and vulnerable to mortgage-rate sensitivity. -
Corporate profits: ~$4.0T → ~$4.3T (Aug 20 to late Aug onward)
This is important: profits stabilized/improved in your sample, aligning with “SAFE profits” today and arguing against near-term recession absent a credit shock. -
Financial conditions (NFCI): ~-0.56 stable through Sep
That’s consistently accommodative. Today’s ~-0.46 is somewhat less loose, but still supportive—consistent with risk score falling over the month rather than rising. -
Credit spreads (HY OAS proxy): ~275 bps → ~260 bps into early Sep
Your older history shows tightening; today’s ~323 bps is wider than that earlier regime. Net: credit is not flashing crisis, but it’s no longer as benign as late summer.
Bottom line across 90 days: the dataset shows no labor-market break, no broad financial tightening shock, but a persistent leading-sector softness (housing pipeline, goods/freight) that keeps recession risk moderate rather than low.
Stock Screener Signals
Today’s flagged names are dominated by “value dividend” screens (ARCC, AIG, BBY, FNF, HMC, T, BCE, LTM) plus a smaller sleeve of oversold growth (CHTR, TLK). That composition generally aligns with a market that is seeking carry (yield) and pricing slower-but-not-recession growth—a classic “late-cycle defensiveness without panic” posture.
Two actionable interpretations:
- The value/dividend clustering suggests investors are more interested in cash-flow durability than maximizing growth multiples. That’s consistent with your broader macro: sentiment is awful, savings cushion is thin, and consumer credit stress is rising, so “getting paid to wait” becomes a dominant positioning style.
- The oversold growth flags (RSI ~28–30) imply pockets of risk appetite are being selectively re-entered after drawdowns—more like mean reversion than a clean new bull leg. In recession setups, oversold screens often expand broadly; here, the list looks more targeted, consistent with your moderate (not high) risk score.
One note: several yields shown (e.g., “1002%”) are mechanically implausible as true dividend yields and likely reflect data quirks (special distributions, annualization artifacts, or screen-field errors). The macro signal you should take is style factor tilt (value/carry vs growth), not the literal yield prints.
Latest Economic Developments
(1) Growth nowcasts remain expansionary, with a small downtick.
Atlanta Fed GDPNow shows Q4 2026 real GDP tracking ~3.4% as of Nov 17, down from ~3.5% on Nov 9. That’s still solid growth momentum into year-end, even if the direction is slightly softer. (atlantafed.org)
(2) October industrial production surprised to the downside.
The Federal Reserve’s G.17 release reported industrial production fell 0.4% in October, after a 0.7% increase in September. That’s not, by itself, recession—but it is consistent with your “freight/goods economy softness” danger signals. (federalreserve.gov)
(3) Retail sales narrative: slowing at the margin, but control group firmer.
AP coverage framed October retail spending as slowing under still-high prices, while noting the “control group” (key for GDP consumption math) rose meaningfully. This combination maps to your thesis: continued expansion with rising volatility around consumption, not a clean collapse. (apnews.com)
(4) Fed messaging: “AI boom supports growth, but inflation risk remains.”
NY Fed’s John Williams emphasized the AI investment surge as a growth driver while warning inflation persistence could keep policy tight for longer. Separate reporting also highlighted Chair Warsh’s concern that an AI disappointment could trigger a sharp equity drawdown and a rise in unemployment—exactly the channel that could quickly move a moderate recession score higher. (axios.com)
(5) Markets: notable equity volatility concentrated in AI leadership.
AP reported a sharp down day on Nov 17 (Dow -1.6%, Nasdaq -0.7%) with continued choppiness in AI-linked names. This supports your “score whipsaws” in the last 10 readings—macro is stable, but risk sentiment is jumpy. (apnews.com)
Near-Term Outlook (Next 30 Days)
Base case for the next month: continued expansion with heightened two-sided risk—a tug-of-war between (a) still-resilient labor/income and (b) leading-sector softness plus inflation/AI-volatility catalysts.
Key near-term catalysts:
- FOMC minutes (released today, Nov 18, 2026) are a potential volatility trigger. The minutes are backward-looking, but investors will parse whether the October hold masked broader support for near-term tightening, especially given the more recent firm inflation prints and Fed commentary. (federalreserve.gov)
- Housing data remains the most recession-sensitive leading cluster in your model; weak prints (permits/starts) can keep the “amber/red” early-cycle warnings lit. (Census shows the new residential construction release cadence and timing.) (census.gov)
- Credit spreads: your stress threshold framing is correct—watch whether HY OAS migrates from ~3.3% toward a materially wider regime; that would turn “late-cycle” into “risk-off macro event.” (equibles.com)
What would move the score meaningfully higher within 30 days:
- A visible break in claims (not a one-week pop—a persistent uptrend),
- Unemployment rising enough to push the Sahm Rule toward 0.50, and/or
- A spread shock plus equity drawdown (especially if concentrated in AI leverage and capex financing).
Long-Term Outlook (3-6 Months)
Three themes dominate the 3–6 month horizon:
1) The U.S. can stay out of recession with “good enough” labor—until it doesn’t.
Your dashboard’s biggest recession-stop sign is still the same: Sahm Rule at 0.00 and claims around 200K. Historically, recessions become “inevitable” only when labor deterioration becomes self-reinforcing (income → spending → layoffs → credit). Right now, that loop has not started.
2) Leading sectors are weakening in a way that raises downside skew.
Housing pipeline weakness (permits/starts), temp-help deterioration, and freight softness matter because they are early-cycle/leading: they often roll over before payrolls do. The risk is not “tomorrow recession,” but “fragility” — meaning a shock (policy, credit, geopolitics, equity drawdown) can tip the system more easily.
3) The “AI capex boom” is both a growth support and a financial stability tail risk.
Fed-linked commentary and reporting increasingly frame AI investment as a driver of activity—but also a possible concentrated vulnerability if financing conditions tighten or expected productivity gains disappoint. In a downside scenario, equity drawdowns plus tighter funding could translate into layoffs, which would then activate your higher-weight labor triggers. (axios.com)
Net: the 90-day trend in your score (down 12 points over 30 days, hovering mid-30s now) suggests recession risk is not accelerating—but the composition implies asymmetric risk: “moderate baseline, but rapid repricing possible.”
What to Watch
Hard thresholds (model-moving):
- Sahm Rule: watch for movement toward 0.50 (trigger zone).
- Initial Claims: a regime shift above ~230K–250K persistently would matter more than any single weekly print.
- HY OAS: watch for sustained widening above the mid-3% area into a higher stress band.
Category watchlist (high signal right now):
- Housing: building permits and starts—any stabilization would reduce tail risk; renewed declines deepen leading weakness.
- Temps / employment mix: temp-help deterioration is already “DANGER”; watch whether it spreads into broader payroll categories.
- Industrial / goods: IP and freight—if industrial production weakness persists alongside negative freight, it raises the probability the slowdown spills into labor.
Event risk:
- FOMC minutes (Nov 18, 2026) tone (tightening bias vs patience). (federalreserve.gov)
- AI-linked equity leadership volatility: a sharp drawdown + spread widening is the quickest path from MODERATE to HIGH risk.
Sources
No data available for this window.