Recession Risk 34/100 — August 21, 2026

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

Today’s Recession Risk Score is 34/100 (MODERATE), and it is unchanged versus 30 days ago. The macro backdrop still reads as “late-cycle but not breaking”: the core recession tripwires tied to unemployment acceleration and credit stress remain untriggered. The main tension is that “soft” deterioration signals (temporary help, sentiment, freight, ultra-low savings) are getting louder even as markets and financial conditions stay broadly supportive. Net: baseline is continued expansion into early fall, with tail risk tied to the Fed’s hawkish reaction function and labor-market follow-through after July’s payroll dip.

Score Trend — Last 30 Days

The score has been range-bound over the last 30 days: Start 34 → End 34 (Δ 0), with a minimum of 34 and a maximum of 44 (average 37). That’s the signature of a system that’s repeatedly flirting with “elevated risk” headlines but then mean-reverting as the hard data refuses to confirm a break.

The shape matters: we’ve seen multiple short spikes (into the high-30s/low-40s) followed by quick resets to 34. That pattern is consistent with event-driven anxiety (Fed minutes, rate volatility, valuation jitters) rather than a durable deterioration in the “real economy” nowcasts. Put differently: investors are nervous about what could happen; the high-frequency and labor-flow data still say it hasn’t happened yet.

Key Drivers

1. Labor-market recession triggers remain untripped (yet hiring momentum is fragile)

   • Initial jobless claims are still low in the latest week (your tracker ~209k; AP notes 206k), consistent with limited layoff stress. (apnews.com)
   • The July jobs report is the offset: nonfarm payrolls -23k; unemployment 4.1% (BLS). (bls.gov)
   • Sahm Rule is still decisively untriggered in your system (SAFE at -0.03), implying no broad-based unemployment shock—this is a critical reason the score is not higher.

2. Yield curve normalization is a near-term “all clear” signal

   • Your 2s10s reading sits at +0.50 (SAFE), aligning with a more standard expansionary configuration rather than the classic inversion signal.

3. Credit stress remains contained

   • High-yield OAS ~271 bps (SAFE) is tight, corroborated by external commentary showing HY spreads in the mid-260s bps in mid-August. (advisorperspectives.com)
   • Tight spreads are inconsistent with an imminent demand shock unless they begin widening sharply (a key tripwire for September).

4. Financial conditions are still loose—supporting risk-taking

   • Chicago Fed NFCI -0.55 (SAFE) plus VIX 14.6 (SAFE) signals broad ease/complacency. But note the VIX is also a “fragility gauge”: low vol can flip fast when the narrative shifts.

5. The Fed is on a hawkish watch—and markets are taking the minutes seriously

   • The Fed held the target range at 3.50%–3.75% on July 29, 2026 (Fed statement). (federalreserve.gov)
   • Minutes released this week reinforced that many officials see a rate-hike scenario if inflation stays elevated, amplifying the probability that September becomes a tightening catalyst. (apnews.com)

6. Household and goods-cycle “soft spots” are real and persistent

   • Personal savings rate 2.7% (DANGER) + UMich sentiment 49.5 (DANGER) = consumers are emotionally and financially stretched even if they’re still employed.
   • Freight index (DANGER) and copper/gold (DANGER) echo the same message: the goods/industrial impulse is weak—often a leading edge before labor cracks.

Category Breakdown

Using your category counts:

• Primary Indicators: 4 safe / 4 watch / 1 danger
  Mixed but not recessionary: unemployment dynamics are “watch,” while the Sahm-style trigger remains safe; the one “danger” is the classic early labor canary (temporary help).

• Secondary Indicators: 2 safe / 0 watch / 1 danger
  The secondary stack is stable on balance, but the “danger” confirms that some cyclicals are already rolling.

• Housing & Construction: 0 safe / 1 watch / 1 danger
  Housing is not collapsing, but it is below trend: starts (WARNING) and permits (WATCH) argue for continued drag rather than a rebound.

• Business Activity: 2 safe / 1 watch / 0 danger
  Business activity is still more “expansion with slowing” than contraction—important ballast for the moderate score.

• Consumer Credit Stress: 1 safe / 2 watch / 1 danger
  Credit is the sleeper risk: delinquencies and debt service are “watch,” and low savings is the accelerant if hiring weakens.

• Market Signals: 7 safe / 2 watch / 5 danger
  Markets are simultaneously “risk-on” (indexes near highs, low vol) and “late-cycle expensive” (valuation-to-GDP extremes, copper/gold fear). This divergence typically raises tail risk rather than immediate recession odds.

• Liquidity: 0 safe / 1 watch / 2 danger
  Liquidity is a growing vulnerability: the ON RRP drawdown/depletion reduces an important buffer and can magnify funding volatility.

• Real-Time / High-Frequency: 0 safe / 1 watch / 1 danger
  High-frequency reads as “slowing but not breaking,” with one danger flag reflecting goods-cycle weakness.

Biggest Movers

Top 5 by absolute 7-day % change (per your block), with interpretation:

1. Bank Unrealized Losses: +931.1% (7D) — Confirmatory (worsening tail risk)
   Even if the level is noisy/definition-shifting, the direction screams sensitivity to rate moves. This is not a recession trigger by itself, but it raises the probability of a liquidity shock channel if yields jump.

2. Conference Board LEI: +673.3% (7D) — Contradictory (improving)
   This magnitude likely reflects a base-effect/data refresh artifact (your history shows occasional jumps). Directionally, a positive LEI supports the moderate score: LEI typically deteriorates meaningfully before recession.

3. Yield Curve (2s30s): +340.0% (7D) — Ambiguous / mildly confirmatory
   A steepening 2s30s can be “good” (term premium rising with growth) or “late-cycle” (market pricing eventual easing). Given hawkish minutes, a steepener here reads more like term-premium/rate-vol than imminent easing.

4. ON RRP Facility: -99.4% (7D) — Confirmatory (worsening liquidity buffer)
   Your data shows RRP collapsing toward effectively depleted levels—this tends to matter most when volatility rises; in calm markets it’s a background vulnerability.

5. VIX: +40.9% (7D) — Confirmatory (risk rising, but not extreme)
   From a low base, a 40% jump can still leave VIX “not high.” But it’s consistent with a market that is starting to price regime risk around Fed policy.

90-Day Indicator Trends

Your “90-day” block (as provided) shows the most recent detailed window clustered from late May through mid-June for many series, but the direction of travel is still informative for a trend read.

Labor & income (the recession kernel)

• Initial claims: 209k (2026-05-23) → 229k (2026-06-12) in your history: a modest drift higher, but still consistent with a healthy labor market. A sustained move toward the mid-200s/high-200s would change the score quickly.
• Sahm Rule: 0.13 → 0.10 (late May to mid-June), drifting lower in that window—consistent with “no unemployment shock.”
• Real personal income ex transfers: $16.7T → $16.5T (late May to early June). That’s a mild negative drift and supports the “slowing momentum” thesis.
• Unemployment rate: your history shows 4.3% in late May/June, while today’s reading is 4.1%—i.e., not a straight line deterioration; labor is softening in hiring, not collapsing in layoffs.

Housing

• Permits: 1442k → 1423k (late May to mid-June), down modestly.
• Starts: largely flat in your history window (~1465k), but today’s dashboard shows 1239k (WARNING), implying housing momentum has worsened since the earlier data points you listed. Housing is one of the more plausible channels through which “higher-for-longer (or re-higher)” bites.

Financial conditions & credit

• NFCI: around -0.52 → -0.51 in your history: still loose.
• HY OAS: mostly low-270s bps in your series, with brief jumps (e.g., 320bps) that then mean-revert. That’s a classic “markets still believe” signal.
• ON RRP: extremely volatile in your history but with a clear tendency toward lower balances, consistent with your current “near depleted” warning.

Markets & valuation

• S&P 500: ~7473 → ~7394 (late May to mid-June in your history), while today is 7786—equities have regained upside and are signaling continued expansion.
• NASDAQ-to-GDP (DANGER): remained extreme throughout your history. That isn’t a recession signal per se, but it does amplify the risk of a “financial accident” if policy tightens unexpectedly.

The key trend takeaway

The 90-day directional read is late-cycle deceleration without the classic recession confirmation: layoffs/claims remain contained and credit spreads remain tight. Where the risk is building is fragility—low savings, depressed sentiment, temp help contraction, and liquidity/bank-duration sensitivity. That combination is why the score stays MODERATE (not low), even with strong equities.

Stock Screener Signals

Your quant flags are dominated by value dividend (ARCC, AIG, BBY, FNF, HMC, T, BCE) plus a couple oversold growth names (CHTR, TLK) and some international cyclicality (LTM, HMC, TLK). The macro read: the market is not positioning for imminent recession via broad defensives; instead it’s leaning into cash-flow/valuation discipline and income, which is typical in a “growth is slower but not negative” regime.

Two notable nuances:

• The oversold growth flags (notably CHTR RSI 28) suggest selective stress under the surface—often tied to balance-sheet sensitivity to rates or sector-specific headwinds. That fits with the “hawkish Fed minutes” theme: duration-sensitive equities can de-rate quickly even when the macro economy is okay.
• The dividend/value clustering reads like a “quality carry” trade—investors seeking equity exposure but with downside buffers (cash returns, lower P/Es). That’s consistent with a moderate recession score: not pricing contraction, but not chasing high-beta cyclicals indiscriminately either.

(One data integrity note for your pipeline: the listed yields—e.g., 1002%—are almost certainly scaling/field issues. Treat the style classification as informative; treat the yield numbers as suspect until normalized.)

Latest Economic Developments

• Weekly jobless claims remain historically low, reinforcing the “no layoff wave” narrative. AP reports claims fell to 206,000 last week, with layoffs still sparse. (apnews.com)
• July payrolls printed negative: the U.S. economy shed 23,000 jobs in July, with unemployment at 4.1%, per the official BLS Employment Situation release. The next Employment Situation release (for August 2026) is scheduled for September 4, 2026—the next major labor catalyst. (bls.gov)
• Fed communications turned more hawkish in tone: the July 29 decision held the policy range at 3.50%–3.75%, but the minutes indicate many officials see a plausible path to hikes if inflation doesn’t cool. This keeps the market focused on the September 16, 2026 FOMC meeting as a live tightening risk. (federalreserve.gov)

In short: the “hard” macro (claims, unemployment level) argues against an imminent recession, but the policy risk premium is rising because the Fed is explicitly keeping hikes on the table.

Near-Term Outlook (Next 30 Days)

Base case for the next month: continued growth with slower momentum, but with higher sensitivity to inflation prints and Fed communication.

Likely score path: 30s unless one of these catalysts hits:

• Labor market follow-through: If claims begin trending meaningfully above the low-200s and the unemployment rate rises enough to push the Sahm Rule toward the 0.50 trigger, the score can jump quickly.
• Credit spread repricing: HY OAS widening from ~270 bps toward ~350–400 bps would be a material warning that financial conditions are tightening endogenously.
• September FOMC repricing: The Fed’s minutes language makes a September hike scenario more plausible; even without a hike, a hawkish dot-path/press conference could tighten conditions.

Known calendar anchor: Employment Situation (August 2026) on September 4, 2026 (BLS schedule), which will be the key “confirmation or refutation” of July’s payroll drop. (bls.gov)

Long-Term Outlook (3-6 Months)

Over a 3–6 month horizon, the economy looks less like a near-term recession setup and more like a late-cycle balance between:

• Supportive financial conditions and tight credit spreads (pro-expansion), versus
• Household fragility + goods-cycle softness + labor leading indicators (temp help) rolling (pro-slowdown).

The most plausible recession pathway into Q4 is not “classic overheating → forced layoffs” yet; it’s policy-induced tightening interacting with fragile buffers:

• If inflation stays sticky and the Fed leans hawkish (or hikes), rate volatility could rise.
• That volatility can transmit through duration-sensitive balance sheets (bank unrealized losses) and through consumer credit (delinquencies rising, savings depleted).
• In that scenario, the recession probability can rise fast even if it starts from a moderate baseline—because the “shock absorbers” are thinner than they look.

Historical parallel (mechanism, not identity): late-cycle periods where labor levels look fine but hiring slows first and then a policy/financial tightening event converts “soft landing” into “hard landing.” The key is whether the labor market transitions from “low layoffs” to “rising layoffs”—claims will tell that story earliest.

What to Watch

High-impact thresholds (score movers):

• Initial claims: sustained break above ~240k–260k would be an early warning that layoffs are broadening. (AP/FRED confirm the current regime is still ~low-200s.) (apnews.com)
• Sahm Rule: movement toward 0.50 is the clean labor-trigger tripwire (your dashboard: -0.03, safely untriggered).
• HY OAS: watch for a regime shift above ~350 bps (tight today; mid-260s to low-270s recently cited). (advisorperspectives.com)
• Fed path: any explicit tilt toward hiking at/after the September 16, 2026 meeting (your stated tripwire) given the minutes’ “hike if sticky inflation” framing. (apnews.com)
• Next jobs report: September 4, 2026 Employment Situation—confirmation risk that July’s -23k was either a one-off or the start of a downshift. (bls.gov)

Sources

No data available for this window.